How Much Does Testosterone Drop During the Day?

By: TRT Provider Guide
Last verified/updated: 2026-09-02

Headline finding: Across 11 male blood-based estimates in this 19-study review, later testosterone was 7.5% to at least 43% lower than the earlier value. The median estimate was 15.0%. The studies used different ages, hours, samples, and designs, so this is a map of published results—not one pooled effect and not a prediction for one person.

How much does testosterone drop during the day in published studies?

Answer: In the 11 male blood-based estimates that allowed a later-versus-earlier comparison, the later result was 7.5% to at least 43% lower. The median estimate was 15.0%. That range combines different study designs. It should not be used as a conversion formula for a late blood draw.

Figure 1. The lowest eligible estimate was 7.5%, the median was 15.0%, and the highest reported lower bound was at least 43%. Source: the 11 eligible male blood estimates in the downloadable study-level CSV. The designs are mixed and are not pooled.

Last verified: September 2, 2026 By TRT Provider Guide · Evidence-cited editorial content; not medically reviewed.

TRT Provider Guide is an independent research and reference resource. It is not a clinic, pharmacy, laboratory, insurer, or medical practice.

The clearest study result: In a 66-man study run during normal clinic hours, total testosterone at 4 p.m. was 20% to 25% lower than at 8 a.m. in men ages 30 to 40. At age 70, the gap was about 10%. And 12 of the 24 men who had at least one afternoon result below 300 ng/dL were at or above 300 ng/dL at all three of their morning visits. — Brambilla et al. (2009), Journal of Clinical Endocrinology & Metabolism; TRT Provider Guide reading of the full text

That is the single most useful published finding on testosterone levels throughout the day. Here is why it is not the whole story.

Across the 19 studies on this page, the reported fall in men's blood testosterone during the day runs from 7.5% to at least 43%. That is a 6.5-fold spread. It is not a fight about biology. It is a difference in what each study compared, who was in it, and how often blood was drawn. Keep that in mind every time a website gives you one number.

One more fact, from a laboratory's own test page. Mayo Clinic Laboratories states that early-morning levels in young men average about 50% higher than afternoon levels. Then it adds a sentence almost nobody repeats:

"Reference values were established using specimens collected in the morning." — Mayo Clinic Laboratories, test catalog entries TGRP and TTST, checked September 2, 2026

In plain words: the "normal range" printed beside your result was built from morning blood.

Bottom line: Testosterone often rises during sleep, is higher after waking, and falls during the waking day. But there is no honest hour-by-hour chart that gives every man one number for every hour. Age, sleep timing, food, health, the lab method, and ordinary day-to-day change all move the result.

Source: The linked primary studies or guideline pages named in the table. Calculated fields follow the methodology on this page.

Original data: Download the 19-study evidence map as CSV. The file records sample size, age, blood or saliva, test times, same-day status, assay notes, whether each percentage was reported or derived, the study's main limit, and identifiers and source links for every row. The Clock Gap Audit is a separate download.

Figure 1. Selected changes reported or derived from the linked studies. Blood and saliva results are kept separate. The bars are not a pooled average and are not a personal prediction.

How to cite this page

Page: TRT Provider Guide. "Testosterone Levels Throughout the Day: What 19 Studies Show." TRT Provider Guide Research. Version 1.1.0. Published and last verified September 2, 2026. https://trtproviderguide.com/research/testosterone-levels-throughout-the-day/

Dataset: TRT Provider Guide. Testosterone Daily Rhythm Evidence Map, 1983–2026. Version 1.1.0. September 2, 2026. https://trtproviderguide.com/research/testosterone-levels-throughout-the-day/#data

If you are quoting one specific number, cite the original study for that number. This page is the comparison across studies; the studies are the source of each figure.

---

How do testosterone levels change throughout the day?

Testosterone does not sit at one level all day. In most men, it rises during sleep, is highest in the early morning, and drifts lower through the waking day. The exact curve is not the same for everyone, and the size of the fall depends heavily on how you measure it.

A 24-hour study of 10 young and 8 middle-aged men drew blood every 30 minutes. Both age groups showed a clear daily rhythm. The fitted peak for total, free, and bioavailable testosterone landed between 7:00 and 7:30 a.m. Every man in the study had at least a 43% fall from his own highest point to his own lowest point. See Diver et al. (2003).

Two details from that study are easy to miss:

• The rhythm was still clearly present in fit men in their sixties. So "the rhythm disappears with age" is too strong. It shrinks.
• SHBG peaks in the afternoon, not the morning. SHBG is the protein that carries most of your testosterone in the blood. Its high point came at about 3:12 p.m. in the young men and 3:40 p.m. in the middle-aged men. That is the opposite direction from testosterone itself, and it is one reason free testosterone and total testosterone do not always move together during the day.

A much larger screening study found a smaller change. Among 3,006 men with a mean age of 60, the average was 444.9 ng/dL in the 6 to 10 a.m. group and 380.4 ng/dL in the 2 to 6 p.m. group. But each man gave only one sample, so this was not a same-person curve. See Crawford et al. (2007).

The lesson is not that one study is right and the other wrong. They measured different things. The European Association of Urology puts it simply in its guideline text: testosterone shows a daily variation that persists into older age, and food intake can lower it too, which is why it asks for fasting morning samples.

A note on units. US labs report testosterone in nanograms per deciliter (ng/dL). Most of the rest of the world uses nanomoles per liter (nmol/L). One nmol/L equals about 28.8 ng/dL. Our testosterone unit converter does the math.

---

How much does testosterone fall from morning to afternoon?

There is no single percentage that fits all men, and the honest answer is a range. Across the studies below, the reported fall in men's blood testosterone runs from 7.5% to at least 43%, with a median of about 14%. The number depends on which two times were compared, how old the men were, and whether the study followed each man across the day or compared different men at different hours.

The Diurnal Drop Ledger

Every published estimate we could find, in one place, sorted from largest to smallest. "Derived" means we calculated the percentage from values printed in the paper; "reported" means the paper states it.

Source: The linked primary studies or guideline pages named in the table. Calculated fields follow the methodology on this page.

Table source and calculation note: TRT Provider Guide, Testosterone Daily Rhythm Evidence Map, 1983–2026, version 1.1.0, verified September 2, 2026. Derived percentages use (morning value − later value) ÷ morning value × 100. For example, Fenn et al. reported 191.2 pmol/L at 9 a.m. and 120.2 pmol/L at 10 p.m.: (191.2 − 120.2) ÷ 191.2 = 37.1%. We did not pool these rows into one average, because the samples, methods, time windows, and designs differ.

Reconciliation note on Crawford 2007. The paper's stated figure of about 13% does not reproduce if you compare the last time band with the first (444.9 to 380.4 is a 14.5% fall). It reproduces exactly, at 13.07%, when the last band is compared with the average of the first three bands (444.9, 433.5, 434.4). We publish the reconciled figure and say which comparison it is, because the difference matters when someone quotes it.

---

Why do some pages say 25% lower while others say 33% or 50% higher?

Both kinds of statement can describe the same gap. They use different starting points, so the denominator changes. A value that is 25% lower in the afternoon is the same as a value that is 33% higher in the morning.

Source: The linked primary studies or guideline pages named in the table. Calculated fields follow the methodology on this page.

This matters in the Brambilla paper. Its abstract says 4 p.m. levels were 20% to 25% lower than 8 a.m. levels in men ages 30 to 40. Its discussion says 8 a.m. levels were about 30% to 35% higher than mid-to-late afternoon levels. Those statements are not a conflict. They use opposite denominators.

It matters for Mayo's number too. "50% higher in the morning" is the same fact as "33% lower in the afternoon." When you see a page quoting 50%, and another quoting 33%, check which direction each one is counting before deciding they disagree.

---

Why do the studies disagree so much?

Four things drive the spread: which two clock times were compared, how old the men were, whether they already had low testosterone, and whether the study measured blood or saliva in the same men or in different men. Change any one of these and the percentage moves.

1. Which two times you compare

Diver's team drew blood every 30 minutes for 24 hours and compared each man's own peak to his own lowest point. That gives 43% or more. But nobody gets a blood test at their personal lowest point. Real appointments happen between about 7 a.m. and 5 p.m. Studies that compare two normal clinic hours get much smaller numbers because they never reach the bottom of the curve.

Crawford's 2015 data shows the shape of this. In 671 men with a mean age of 61, the 11 a.m. to 2 p.m. window averaged 396.5 ng/dL, close to the 8 to 11 a.m. average of 411.7 ng/dL. The sharper fall came later: 344.3 ng/dL at 2 to 5 p.m. and 334.4 ng/dL at 5 to 8 p.m. So the picture is less "steady decline all day" and more "a high plateau through the morning, then a fall in the afternoon."

2. How old the men were

This is the biggest single reason the numbers split, and one study proves it cleanly. Brambilla's team used the same method, the same clock times, and the same group of men. The only thing that changed was age.

Source: The linked primary studies or guideline pages named in the table. Calculated fields follow the methodology on this page.

Source: Brambilla et al. (2009). The 2.25-fold ratio uses the midpoint of the 20% to 25% band and is our arithmetic.

That is roughly a 2.25-fold difference from age alone, with everything else held still. Welliver's team saw the same pattern in 2,569 men over 15 years:

Source: The linked primary studies or guideline pages named in the table. Calculated fields follow the methodology on this page.

Source: Welliver et al. (2014), Journal of Urology, via the published J Fam Pract summary. The authors judged the 70–74 gap real but unlikely to change any clinical decision.

3. Whether the men were already low

Shlykova's team took young men and split them into two groups, normal and low, then drew blood six times over 24 hours. The normal group fell 15% between 8 a.m. and 2 p.m. The low group showed no significant daily rhythm at all. Winters found the same thing in 1991: the men with primary testicular failure showed no significant rhythm.

This is worth sitting with. The daily cycle is often absent in exactly the men who are being tested for low testosterone.

4. Blood vs. saliva, same men vs. different men

Saliva testosterone is a different measurement from blood testosterone and moves on a different scale. And a study that compares the same men at two times of day answers a different question from a study that compares different men who happened to arrive at different hours. The evidence map labels every row on both points so they are never confused.

---

Does age change the daily testosterone pattern?

The daily swing tends to get smaller with age, but there is no sound one-size-fits-all age correction. Older men can still have a lower afternoon result, and every guideline still asks for a morning test.

The oldest study on this page, Bremner et al. (1983), took hourly blood samples for 24 hours in 17 young men (mean age 25) and 12 older men (mean age 71). The early-morning rise seen in the young group was much weaker or absent in the older group. Brambilla's clinic-hours model found a 20% to 25% gap at ages 30 to 40 and about a 10% gap at age 70. Welliver found 207 ng/dL under age 40 and 34 ng/dL at 70 to 74.

In men 45 and older, Guay et al. (2008) found that total testosterone did not differ between an 8 to 10 a.m. draw and a 10 a.m. to noon draw, though free and bioavailable testosterone were higher at the earlier time. Note that this study compared two morning windows, not morning against afternoon.

Newer work adds caution. Novaes et al. (2024) studied 506 men with a median age of 65. The link between age and the size of the daily change was real but weak. The biggest changes were in the small group under age 30. A 2025 saliva study of 104 Shuar males in Ecuador also found less morning-to-evening variation with age, but it covered one population and included boys as young as 12. See Gildner et al. (2025).

The safe summary: age often blunts the pattern, but it does not erase the need for correct test timing.

For what counts as a typical level at each age, see our testosterone levels by age chart. That page covers the reference range question; this page covers the hour.

---

Is the testosterone peak tied to clock time or to sleep?

Sleep drives a large part of the pattern. The rise is not controlled by the wall clock alone, which is why a fixed morning window can miss for night-shift workers.

In a controlled study of seven healthy men, Axelsson et al. (2005) moved sleep from night to daytime. Testosterone rose during both night sleep and shifted day sleep, then fell during waking. In another study of ten men, breaking sleep into short sleep-and-wake blocks delayed and weakened the usual overnight rise. See Luboshitzky et al. (2001).

Short sleep lowers the daytime level too. Leproult and Van Cauter (2011) limited ten healthy young men to about five hours of sleep a night for a week. Daytime testosterone fell roughly 10% to 15%, with the lowest readings between 2 p.m. and 10 p.m. That is the same order of magnitude as the daily rhythm itself.

According to the current EAU male hypogonadism guideline, the standard clock window may not fit night-shift workers, who should be sampled while fasting and after an adequate sleep period. We took that wording from a reading of the current guideline page; readers should verify it against the EAU text before quoting it.

---

How much can the hour of your appointment change the verdict?

Using one identical morning result and each published drop in turn, the afternoon number lands anywhere from 200 to 327 ng/dL, and 7 of the 12 estimates push it below the 300 ng/dL line. This is not a conversion tool. It is a way to see how far apart the published estimates sit.

Take one man with an 8 a.m. result of 350 ng/dL, a normal result at every major US lab. Now apply each published drop for men's blood testosterone:

Source: The linked primary studies or guideline pages named in the table. Calculated fields follow the methodology on this page.

Source: TRT Provider Guide arithmetic, verified September 2, 2026. The 300 ng/dL line is the American Urological Association's threshold supporting a diagnosis of testosterone deficiency. The 264 ng/dL figure is the bottom of Labcorp's printed adult male reference interval. See our TRT bloodwork reference ranges for how labs and guidelines differ.

One identical morning result. Twelve published estimates. A 127-point spread in the afternoon number, and a coin flip on the diagnosis. That is exactly why no single multiplier can be trusted.

Does the hour actually change who gets diagnosed?

The evidence is genuinely split, and this is the most important honest caveat on the page.

Two large studies found the hour barely moved the diagnosis rate:

• Crawford (2007), 3,006 men across four time bands from 6 a.m. to 6 p.m. Average testosterone fell in the last band. But the share of men under 300 ng/dL did not change across the four bands.
• Novaes (2024), 506 men with paired morning and afternoon draws measured by mass spectrometry. 63% had no meaningful change. 24% dropped by more than 100 ng/dL. And 13% went up by more than 100 ng/dL. That last number is rarely mentioned anywhere. "Afternoon always reads lower" is not true for one man in eight.

Two other findings point the other way:

• Brambilla (2009): of 24 men with at least one afternoon result under 300 ng/dL, 12 were at or above 300 at all three morning visits. For those men, the afternoon draw would have been the misleading one.
• Welliver (2014): a 207 ng/dL gap in men under 40 is large enough to move almost anyone across a diagnostic line.

Our reading, stated as a judgment and not a fact: the hour matters most for younger men and least for older men. That is the one pattern every study here agrees on, even when they disagree about the size. It is also why the guidelines all still say morning: the rule is cheap to follow and occasionally prevents a real error.

---

Does the lab tell you any of this?

Mostly no. Of seven sources a man actually encounters, six tell him when to get blood drawn, one puts a number on how much the result moves, and one tells him the normal range itself came from morning samples. We opened each source on September 2, 2026 and recorded what it says.

The Clock Gap Audit

Source: The linked primary studies or guideline pages named in the table. Calculated fields follow the methodology on this page.

Source: TRT Provider Guide, Clock Gap Audit, all seven sources opened and verified September 2, 2026.

The Labcorp row is the one we did not expect. Test 004226 is one of the most commonly ordered testosterone tests in the United States. On Labcorp's own test page, the collection instructions cover which tube to use. There is no collection time on the page, and no mention of the daily rhythm anywhere on it.

We are not saying Labcorp is wrong. A lab test page is written for the clinician ordering the test, and the ordering doctor is expected to know the timing rule. But a growing number of men order these tests themselves, with no doctor in between. For them, that page is the only instruction they get.

The guidelines are clear about the rule and quiet about the number. The AUA says the diagnosis should be made only after two total testosterone measurements on separate occasions, "both conducted in an early morning fashion." The EAU names a hard window of 7 to 11 a.m. with fasting. None of them tells you how much your result moves if you miss the window. This page exists to fill that gap.

---

Is the clock even the biggest thing that moves a same-day result?

No. For a man in his 30s, the hour and a sugary drink move the number by about the same amount. For a man of 70, the drink matters roughly 2.5 times more than the hour. The clock gets all the attention, but it competes with things that are just as large and much easier to control.

Source: The linked primary studies or guideline pages named in the table. Calculated fields follow the methodology on this page.

Source: TRT Provider Guide comparison of published figures, verified September 2, 2026. The 2.5-fold ratio is our arithmetic (25 ÷ 10).

The glucose finding is striking. Caronia's team gave 74 men a standard sugary drink, the same one used in diabetes testing, and mean testosterone fell about 25% within two hours. That is bigger than the clock effect for most men in this table.

The British Society for Sexual Medicine says the same thing in its 2023 guideline: measure ideally fasting, before 11 a.m., and expect non-fasting levels to run up to 30% lower. This is why every guideline says fasting morning blood, not just morning blood. The two instructions are doing different jobs.

---

What time should a testosterone blood test be done?

For a normal sleep schedule, the best-supported diagnostic window is a fasting morning sample, repeated on another day if the first result is low. Every major body agrees on that, and the EAU gives the clock time: 7 to 11 a.m.

Source: The linked primary studies or guideline pages named in the table. Calculated fields follow the methodology on this page.

Source: TRT Provider Guide reading of each guideline, verified September 2, 2026. Our TRT guidelines comparison covers the full documents.

The Endocrine Society's 2018 guideline explains the reasoning in one paragraph: testosterone peaks in the morning, aging reduces the size of that peak, but a substantial fraction of men aged 65 to 80 who read low in the afternoon will read normal in the morning, and food and glucose lower the level. So: morning, fasting.

A simple testing checklist:

1. Follow the instructions from the lab or clinician who ordered the test.
2. For a normal sleep schedule, test in the morning, before 11 a.m.
3. Fast when the order or guideline calls for it. Skip the sugary drink.
4. After a low result, repeat the test on another morning before drawing any conclusion.
5. When tracking change over time, use the same lab and a similar hour.
6. Do not diagnose low testosterone from one number alone.

This page is educational. A qualified clinician should read a result alongside symptoms, health history, medicines, sleep, and the lab's method and range.

---

Do men with low testosterone still rise and fall the same way?

Not always. Small studies suggest the daily swing can be weaker or hard to detect in some low-testosterone groups, which is one more reason a universal correction formula is not safe.

In Winters et al. (1991), the five men with primary testicular failure were 11.5% lower at 4 p.m., but the study did not find a significant daily rhythm in that group. The five men with high prolactin were 35% lower. In Shlykova et al. (2020), the normal group fell 15% from 8 a.m. to 2 p.m., while the low morning group fell about 4% and the change was not significant. A model-based paper, Gupta et al. (2000), also did not detect a rhythm in its hypogonadal profile.

These samples were small or model-based. They do not prove that a man with low testosterone has no daily change. They show why one formula cannot fit everyone.

---

Do women's testosterone levels change during the day?

Women may also have a daily change, but the evidence is thin and the clinical rules are not the same as for men. The evidence map includes one direct two-time-point blood study in women.

Algburi et al. (2023) tested 46 women ages 21 to 40 at 8:30 a.m. and 1:30 p.m. Mean total testosterone was 23.4 ng/dL in the morning and 21.7 ng/dL in the afternoon, a derived 7.3% fall, and the difference was statistically significant. The study came from one center, used only two times, and did not require fasting. Male cutoffs and male testing claims should not be copied onto women.

---

Does the same daily pattern apply when someone uses TRT?

No. This evidence map is about natural testosterone production in people not using testosterone therapy. It should not be used to adjust a TRT dose or to "correct" a treated patient's result.

The Brambilla paper itself says its findings do not necessarily apply to patients on testosterone replacement when clinicians make dose decisions. TRT products create their own peak-and-trough pattern based on the product, the dose, the route, and the time since the last dose. For a treated patient, the right blood-draw time depends on the treatment plan and the product's monitoring guidance. Our page on what TRT is covers the basics of how the products differ.

---

What do the 19 studies show?

The Testosterone Daily Rhythm Evidence Map is an original structured review of 19 direct human studies published from 1983 through the January 2026 issue of Annals of Clinical Biochemistry. It aggregates published evidence. It is not a new clinical trial and not a pooled meta-analysis.

Source: The linked primary studies or guideline pages named in the table. Calculated fields follow the methodology on this page.

Table source: TRT Provider Guide, Testosterone Daily Rhythm Evidence Map, 1983–2026, version 1.1.0. Each linked study name goes to the original publisher or PubMed/PMC record. Two rows without links (Welliver 2014, Crawford 2015) were extracted from published summaries; their full-text identifiers are listed in the dataset file for confirmation.

---

What does this evidence show, and what does it not?

The evidence shows a real daily pattern that is larger in younger men, shaped by sleep, and sensitive to study design. It does not provide a universal hourly chart, a single percentage, or a way to convert one time of day into another.

The evidence supports these points:

• Testosterone usually falls during the waking day.
• The fall is larger in younger men than in older men.
• Sleep timing can move the rise, and short sleep can lower the whole day.
• Study design changes the percentage: peak-to-low, clinic-hours, same-day, different-day, blood, and saliva are different measures.
• An afternoon result can be lower than repeated morning results, and, less often, higher.
• Food and sugar lower the level by as much as the clock does.
• Morning, fasting, repeated testing is the safer diagnostic approach.

The evidence does not provide these things:

• one normal ng/dL number for every hour;
• one percentage that applies to every man;
• a valid way to convert an afternoon result into a morning result;
• a diagnosis based on time of day alone;
• a dosing rule for people using TRT.

---

How was the Testosterone Daily Rhythm Evidence Map built?

The map asks what direct human studies have found about testosterone levels throughout the day, and it keeps unlike designs separate instead of forcing them into one average. Every derived number on this page can be recalculated from the values printed in the source.

Inclusion rules

A study was eligible when it did at least one of these things:

• measured testosterone at two or more times within a day;
• analyzed time of day as the main exposure;
• tested how sleep timing changed the daily pattern;
• compared morning and later results in a clinically relevant group.

We focused on human data. We kept blood and saliva studies separate. We included male studies, one direct female blood study, natural-sleep studies, sleep-shift studies, and clinic-hours morning-versus-afternoon studies, including the large screening and database cohorts.

What we extracted

For each study: publication year and source; sample size and age; sex and population; blood or saliva; the stated analyte; sample times and frequency; same-day or different-day design; sleep and food controls when reported; assay when reported; the study's stated result; any percentage we derived from printed values; the main limit that affects interpretation; and DOI, PMID, PMCID, source link, and verification date.

How derived percentages were calculated

When a paper printed a morning value and a later value but not the percentage, we used:

(morning value − later value) ÷ morning value × 100

Derived values are labeled derived on the page and in the CSV. Reported values stay labeled reported. We did not turn a ratio, a curve amplitude, a share of decline, or a different-day change into a fixed-time percentage unless the calculation matched the paper's stated values and meaning.

Why there is no pooled "average daily drop"

A pooled average would look simple but would mix unlike evidence. A 24-hour personal peak-to-low value is not the same as an 8 a.m. to 4 p.m. value. Saliva is not blood. A same-person test is not the same as comparing different men. A same-day comparison is not the same as one taken up to three months later. We show the findings side by side and label the design instead.

Three source problems we caught, published here because they change the numbers

1. Brambilla's abstract and its Table 2 disagree. The abstract reports 17 men with a low afternoon result who were normal every morning. The full-text Table 2 and discussion report 12 of 24, or 50%. We use the table-backed figure because its age-group cells can be checked and sum to all 62 men with complete data. The mismatch is preserved in the claim ledger.
2. Crawford 2007's "about 13%" is the last band against the mean of the first three, not against the first band alone. The two calculations give 13.07% and 14.50%. We say which one we publish.
3. Shlykova's low-testosterone group is a null result, not a 4% drop. The published means divide to about 4%, but the authors report no significant rhythm. We show the number, label it not significant, and exclude it from the range.

Search and verification process

The source set was built from PubMed and PMC records, original publisher pages, the reference lists of the primary papers, laboratory test catalogs, and current pages from the American Urological Association, Endocrine Society, European Association of Urology, and British Society for Sexual Medicine. Each included record was checked on September 2, 2026. Two studies (Welliver 2014 and Crawford 2015) were extracted from published clinical summaries rather than the full papers, and the table says so.

This is a focused evidence map, not a registered systematic review. It may not contain every paper ever published on testosterone timing. Its value is the row-level normalization, the source checking, the design labels, and the open data.

Version rules

• Version 1.0.0: first published evidence map, 15 studies, verified September 2, 2026.
• Version 1.1.0: added four clinic-hours cohorts (Guay 2008, Welliver 2014, Crawford 2015, Fantus 2024), the Clock Gap Audit, and the same-day factor comparison. 19 studies.
• A major version changes if the inclusion method changes. A minor version changes when studies or fields are added. A patch version fixes an extraction or file issue.
• The visible "last verified" date changes only after the sources are checked again.

---

What are the limits of this data?

The biggest limit is that the studies do not all measure the same thing in the same way. They should be compared by design, not blended into one number. We would rather say the rest of this plainly than have a careful reader find it.

• These are group averages, not personal predictions. The verdict-flip table shows the spread across published estimates. It does not tell any man what his own afternoon number would be.
• Sample sizes are wildly uneven. Diver's 43% comes from 18 men. The UK Biobank's 14.4% comes from 197,883. They should not carry equal weight, and the map shows both so you can weigh them.
• Most of the clinic-hours studies used men who came in for a testosterone problem. Crawford, Welliver, and Novaes all drew from clinical populations. Healthy men might behave differently.
• Only Brambilla holds the method constant across ages. It is the one fair within-study age comparison available, and it is 66 men. Every other age comparison here is across studies, which is weaker.
• Several older abstracts do not report every method detail, and clock time and time since waking are not always both recorded.
• Meals, illness, obesity, medicines, and sleep are not controlled in every study.
• "Statistically significant" does not always mean large or clinically important. Welliver's 34 ng/dL gap at ages 70 to 74 is the clearest example.
• The Mayo figure is laboratory guidance, not a study. We include it because men are given it, and we label it.
• Two rows were built from published summaries, not full papers. They are marked.
• Clinical guidelines change. The five bodies in the timing table should be rechecked before any of their wording is quoted.
• This is educational research, not medical advice. No page can tell you whether your own result is normal. That is a conversation for a clinician who can see your full picture.

---

Frequently asked questions

These short answers cover the most common follow-up questions about daily testosterone change and blood-test timing. They use the same 19-study evidence map and the guideline sources described above.

What time of day is testosterone highest?

In men who sleep at night, early morning. The densest 24-hour study put the peak between 7:00 and 7:30 a.m. Levels stay relatively high through the morning and fall more sharply in the afternoon. Shift work can move the pattern.

How much does testosterone drop by evening?

There is no one answer. Clinic-hours blood studies in men found group values about 10% to 25% lower later in the day, with the larger figures in younger men. A 24-hour peak-to-low study found at least 43% in every subject. Those are different measures, and the full spread across published estimates is 7.5% to at least 43%.

Is testosterone always lower in the afternoon?

No. The group average usually goes down, but one man's later result can stay close to the morning value or even go up. In the newest paired study, 13% of men were more than 100 ng/dL higher in the afternoon.

Should I get my testosterone tested in the morning?

Every major guideline says yes. The AUA asks for two early-morning measurements on separate days. The EAU names a window of 7 to 11 a.m., fasting. The Endocrine Society asks for a fasting morning sample, repeated. The rule is easy to follow and removes one source of doubt from the result.

Does the time of day matter less as you get older?

Usually, yes. Brambilla found 20% to 25% at ages 30 to 40 and 10% at age 70 using identical methods. Welliver found a 207 ng/dL gap under age 40 and a 34 ng/dL gap at 70 to 74. Whether that means an older man can safely test in the afternoon is a clinical judgment, and the guidelines have not changed their advice.

Can an afternoon test show a falsely low result?

It can. Brambilla found 12 of 24 men with a low afternoon result who were normal at all three morning visits. But the reverse is also documented, so a single reading at any hour is weak evidence. That is why the guidelines ask for two.

Why does my lab report show a "normal" range that ignores the time of day?

Because the range was built from morning samples. Mayo Clinic Laboratories states this openly on its test pages. Most lab pages do not mention it. The range is a fair comparison for morning blood and a looser one for afternoon blood.

Can I multiply an afternoon result to estimate my morning level?

No. There is no verified multiplier. The published drops for men's blood testosterone run from 7.5% to at least 43%, and applying them to the same result gives afternoon values from 200 to 327 ng/dL. A single correction factor would be false precision.

Does fasting matter as much as the time?

For many men, more. Caronia's team found mean testosterone fell about 25% after a standard sugary drink. That is as large as the clock effect for a man in his 30s and larger than it for a man of 70. The BSSM guideline says non-fasting levels may run up to 30% lower.

Does poor sleep lower testosterone?

Yes, measurably. A week of about five hours of sleep a night lowered daytime testosterone roughly 10% to 15% in healthy young men. That is the same order of magnitude as the daily rhythm itself.

What time should a night-shift worker test?

According to the EAU guideline, the usual clock window may not fit night-shift workers, and the sample should be taken fasting after an adequate sleep period. Testosterone tracks sleep more than the wall clock.

Does TRT follow the same morning-to-evening pattern?

Not in a simple way. Treatment creates a product- and dose-timing pattern of its own, so monitoring should use the time set by the prescriber or the product guidance. This page does not apply to treated patients.

Do women's testosterone levels change during the day?

They may. One 46-woman blood study found a 7.3% derived fall from 8:30 a.m. to 1:30 p.m., but female testing cannot be interpreted with male cutoffs or male rules.

---

What the 7.5% to at least 43% range means

The range uses 11 male blood-based estimates from the 19-study evidence map. Each estimate had an earlier and later value that could be compared without inventing missing data. For each eligible estimate, the later-versus-earlier difference was calculated as (earlier value − later value) ÷ earlier value × 100. Values were rounded to one decimal place with ordinary half-up rounding. The median is the middle estimate after sorting the 11 percentages.

The 11 values were not pooled. A same-man 24-hour peak-to-low comparison, a fixed morning-versus-afternoon clinic comparison, and a population study based on one draw per person do not answer the same question. They remain separate in the dataset and table. The Novaes medians of 387 and 373 ng/dL are shown as reported, but no percent drop is calculated from them because those medians do not establish a paired within-man decline and the paper reported no significant mean difference.

Download the study data

• Study-level CSV
• Reproduction script
• Chart as SVG
• Chart as PNG

These files reproduce the published estimates and make every included and excluded calculation visible.

AUA fasting note: The AUA requires two early-morning total testosterone measurements on separate occasions. Its diagnostic recommendation does not state that those samples must be fasting. The Endocrine Society, EAU, and BSSM wording is shown separately in the guideline table.

Interpretation rule: The 11 percentages are not pooled. Same-man 24-hour studies, fixed clinic-hour comparisons, and one-draw population studies answer different questions. The table keeps those designs separate, and the range is not a conversion rule for one person.

How to cite this page

Suggested citation: TRT Provider Guide. “How Much Does Testosterone Drop During the Day? 2026 Data.” TRT Provider Guide. Last verified September 2, 2026.

This citation is attribution information only. The study table and downloadable CSV contain the underlying source map and calculations.

Primary sources

The sources below are the 19 direct human studies, the laboratory and patient-education pages in the Clock Gap Audit, the two same-day factor studies, and the issuing-body guidance used on this page. Each link goes to the original publisher, PubMed or PMC record, or issuing organization. Entries without a link are cited by journal and were extracted from published summaries; identifiers should be confirmed before quoting.

Direct human studies

1. Bremner WJ, Vitiello MV, Prinz PN. Loss of circadian rhythmicity in blood testosterone levels with aging in normal men. J Clin Endocrinol Metab 1983;56(6):1278–1281.
2. Winters SJ, et al. Diurnal rhythm of testosterone and luteinizing hormone in hypogonadal men. J Androl 1991;12(3):185–190.
3. Gupta SK, Lindemulder EA, Sathyan G. Modeling of circadian testosterone in healthy men and hypogonadal men. J Clin Pharmacol 2000;40(7):731–738.
4. Luboshitzky R, et al. Disruption of the nocturnal testosterone rhythm by sleep fragmentation in normal men. J Clin Endocrinol Metab 2001;86(3):1134–1139.
5. Diver MJ, Imtiaz KE, Ahmad AM, Vora JP, Fraser WD. Diurnal rhythms of serum total, free and bioavailable testosterone and of SHBG in middle-aged men compared with those in young men. Clin Endocrinol (Oxf) 2003;58(6):710–717.
6. Axelsson J, et al. Effects of acutely displaced sleep on testosterone. J Clin Endocrinol Metab 2005;90(8):4530–4535.
7. Crawford ED, Barqawi AB, O'Donnell C, Morgentaler A. The association of time of day and serum testosterone concentration in a large screening population. BJU Int 2007;100(3):509–513.
8. Guay A, Miller MG, McWhirter CL. Does early morning versus late morning draw time influence apparent testosterone concentration in men aged ≥45 years? Data from the Hypogonadism In Males study. Int J Impot Res 2008;20(2):162–167.
9. Brambilla DJ, Matsumoto AM, Araujo AB, McKinlay JB. The effect of diurnal variation on clinical measurement of serum testosterone and other sex hormone levels in men. J Clin Endocrinol Metab 2009;94(3):907–913.
10. Panizzon MS, et al. Genetic and environmental influences of daily and intra-individual variation in testosterone levels in middle-aged men. Psychoneuroendocrinology 2013.
11. Welliver RC Jr, Wiser HJ, Brannigan RE, Feia K, Monga M, Köhler TS. Validity of midday total testosterone levels in older men with erectile dysfunction. J Urol 2014;192(1):165–169. Summarized in Long N, Nguyen L, Stevermer J. It's time to reconsider early-morning testosterone tests. J Fam Pract 2015;64(7):418–419.
12. van Kerkhof LWM, et al. Diurnal variation of hormonal and lipid biomarkers in a molecular epidemiology-like setting. PLoS One 2015;10(8):e0135652.
13. Crawford ED, et al. Effect of time of day on serum testosterone concentration. Curr Med Res Opin 2015;31(10):1911–1914.
14. Shlykova N, et al. Absent diurnal variation in serum testosterone in young men with testosterone deficiency. J Urol 2020;203(4):817–823.
15. Algburi HD Sr, et al. Diurnal variation of serum total testosterone in women: a single-center study from Basrah. Cureus 2023;15(10):e47677.
16. Novaes LF, Flores JM, Benfante N, Schofield E, Katz DJ, Nelson CJ, Mulhall JP. Analysis of diurnal variation in serum testosterone levels in men with symptoms of testosterone deficiency. J Sex Med 2024;21(5):408–413.
17. Fantus RJ, et al. Novel reference range values for serum testosterone: a cross-sectional study of 200,000 males. J Endocrinol Invest 2024.
18. Gildner TE, et al. Salivary testosterone, age, and adiposity associations among Shuar males in Amazonian Ecuador challenge assumptions of "normal" testosterone patterns. Am J Hum Biol 2025.
19. Fenn J, et al. Diurnal variation in salivary testosterone independent of food consumption. Ann Clin Biochem 2026 (online June 2025).

Same-day factors

20. Caronia LM, Dwyer AA, Hayden D, Amati F, Pitteloud N, Hayes FJ. Abrupt decrease in serum testosterone levels after an oral glucose load in men. Clin Endocrinol (Oxf) 2013;78(2):291–296.
21. Leproult R, Van Cauter E. Effect of 1 week of sleep restriction on testosterone levels in young healthy men. JAMA 2011;305(21):2173–2174.

Laboratory and patient-education pages

22. Mayo Clinic Laboratories. Testosterone, Total and Free, Serum (TGRP). Also Testosterone, Total, Mass Spectrometry, Serum (TTST). Checked September 2, 2026.
23. Labcorp. Testosterone, Total, test 004226. Checked September 2, 2026.
24. MedlinePlus, National Library of Medicine. Testosterone Levels Test. Checked September 2, 2026.
25. University of Iowa Health Care, Laboratory Test Handbook. Testosterone, Free and Total, Adult. Checked September 2, 2026.

Guidelines and statements

26. Mulhall JP, et al. Evaluation and management of testosterone deficiency: AUA guideline. J Urol 2018;200(2):423–432.
27. Bhasin S, Brito JP, Cunningham GR, et al. Testosterone therapy in men with hypogonadism: an Endocrine Society clinical practice guideline. J Clin Endocrinol Metab 2018;103(5):1715–1744.
28. European Association of Urology. EAU Guidelines on Sexual and Reproductive Health, male hypogonadism chapter.
29. Hackett G, et al. British Society for Sexual Medicine guidelines on male adult testosterone deficiency, with statements for practice. World J Mens Health 2023;41(3):508–537.
30. Endocrine Society. Statement on Testosterone Replacement Therapy. July 16, 2026.

---

Editorial identity: TRT Provider Guide is an independent research and reference resource. It is not a clinic, pharmacy, laboratory, insurer, or medical practice. This page carries no advertising, no product recommendations, and no affiliate links. Purpose: Evidence over hype. Clarity over confusion.

Last verified: September 2, 2026. Next scheduled re-check: December 2026 (guideline and lab pages), March 2027 (full dataset).