Coffee and Testosterone: Does It Raise or Lower It?

“Does coffee raise or lower testosterone” sounds like it should have one answer, and a 2026 study of 2,264 Finnish adults shows exactly why it does not. In the same dataset, from the same people, more coffee tracked with more total testosterone in men and, at the same time, with less of the testosterone their bodies could actually use. That is not two studies disagreeing. It is one real hormonal signature with two faces, the same shape this site’s own coffee and cortisol breakdown found for caffeine’s effect on a different hormone: a short-term and long-term picture that both look true because they are measuring different things. This page walks through what the new testosterone data actually found, why a single cup can move one number up and a related number down at the same time, and what a separate, much smaller study found happens on the much faster timescale of a single pre-workout cup.

The study: 2,264 Finnish adults, one hormone panel, two directions at once

The data comes from the Northern Finland Birth Cohort 1966 (NFBC1966), a long-running research cohort that has followed everyone born in the Oulu and Lapland regions of Finland in 1966. Researchers drew on blood work and coffee-intake data collected when cohort members were assessed at age 46, covering 2,264 participants, 2,194 of whom reported drinking coffee and 70 of whom did not, with roughly 47% of the sample men. The findings were published in 2026 in the European Journal of Nutrition, and independently reported by both News-Medical and the University of Oulu’s own newsroom, which confirm the same sample size, cohort, and core numbers.

The researchers measured four related but distinct hormone values for each participant: total testosterone, bioavailable testosterone, sex hormone-binding globulin (SHBG), free testosterone, and a calculated free androgen index (FAI). That list matters more than it looks, because the study’s central finding only makes sense once those four numbers are treated as separate questions rather than one.

What higher coffee intake did to men’s numbers

For men in the cohort, each additional cup of coffee consumed per day was associated with the following average changes, all measured against the rest of the same hormone panel in the same people:

Read left to right, that table tells two opposite stories. Total testosterone and bioavailable testosterone both went up with more coffee. SHBG went up too, by a larger relative step. And free testosterone and the free androgen index, the two measures that most directly reflect the testosterone actually available for the body to use, both went down. High coffee consumers in the cohort showed significantly higher total testosterone than non-consumers and low consumers, so this is not a marginal, barely-there effect sitting at the edge of statistical noise. It is a clear, directionally consistent pattern that simply points two different ways depending on which of the four numbers someone reads.

What higher coffee intake did to women’s numbers

The pattern in women, after excluding participants with polycystic ovary syndrome (PCOS) from the analysis, was narrower but told a consistent part of the same story. Each additional cup of coffee per day was associated with SHBG rising by 1.27 nmol/L, more than double the per-cup SHBG increase seen in men, alongside a 0.32 drop in the free androgen index and a 0.01 nmol/L drop in free testosterone. Total and bioavailable testosterone showed no significant association with coffee intake in women at all. The University of Oulu’s own summary of the finding puts it plainly: women’s “hormonal associations were more limited and were primarily characterised by higher SHBG and lower measures of free androgens.” In other words, women in this cohort got the SHBG-driven, usable-testosterone-down half of the pattern without the total-testosterone-up half that showed up in men.

Why one number goes up while a related number goes down: the SHBG mechanism

The reason total testosterone and free testosterone can move in opposite directions at the same time comes down to how testosterone actually travels through the bloodstream. Sex hormone-binding globulin is the body’s main transport protein for testosterone, made mostly in the liver, and according to Mayo Clinic Laboratories’ own overview of these tests, most of the testosterone circulating in blood is bound to SHBG, with a smaller share bound to albumin and only a small fraction floating free. SHBG “transports testosterone and other steroids in the blood plasma, reduces their metabolic clearance rate, and regulates their access to target tissues,” meaning testosterone attached to SHBG is not available for tissues to actually use the way free or albumin-bound testosterone is.

That mechanism explains the entire pattern in the table above. If coffee raises SHBG, more of the testosterone circulating in the blood gets bound up and taken out of immediate circulation, even while the total amount of testosterone the body is producing, and therefore the total testosterone measurement, goes up at the same time. Total testosterone measures everything in the blood, bound and unbound alike. Free testosterone and the free androgen index are closer approximations of what tissues can actually draw on. A rising SHBG number can push the first kind of measurement up and the second kind down simultaneously, in the same blood draw, from the same person. Mayo Clinic Laboratories is explicit that this is exactly the scenario where relying on total testosterone alone can mislead: “a number of conditions and medications can increase or decrease SHBG concentration, which may cause total testosterone concentration to change without necessarily influencing bioavailable or free testosterone concentration,” which is precisely the split this study found with coffee specifically.

A bonus finding buried in the same dataset: body composition

The same NFBC1966 analysis measured more than hormones, and one additional finding is worth noting because it adds useful context rather than more confusion. Despite having a similar body mass index (BMI) to lower coffee consumers, higher coffee consumers in the cohort showed lower total and visceral fat, greater skeletal muscle mass, and lower levels of branched-chain amino acids, a metabolic marker often elevated alongside insulin resistance. That pattern does not prove coffee changes body composition any more than it proves coffee changes testosterone. It is the same kind of observational association running alongside the hormone findings, worth knowing about, but carrying the exact same causality caveat covered in the next section.

The acute flip side: what one pre-workout dose does in an hour, not a decade

Everything above comes from habitual, long-term coffee drinking measured across years in a single cross-sectional snapshot. A completely separate, much smaller study shows that a single large caffeine dose produces a strikingly similar shape of result on a timescale of about an hour. In a 2008 double-blind, crossover trial published in the International Journal of Sport Nutrition and Exercise Metabolism, researchers led by C. Martyn Beaven gave 24 professional rugby-league players caffeine doses of 0, 200, 400, and 800mg, in random order, one hour before a resistance-exercise session, with saliva sampled throughout to track testosterone and cortisol.

Resistance exercise alone raised testosterone by about 15%. Caffeine added a further, dose-dependent boost on top of that, up to a 21% increase at the highest 800mg dose, roughly equivalent to this site’s own 96mg-per-cup filter coffee figure poured about 8 times, or 13 63mg espresso shots, in a single sitting. That sounds like an unambiguous testosterone win. But the same 800mg dose also produced a 52% increase in cortisol, a stress hormone that works against testosterone’s anabolic, muscle-building effects. Because cortisol rose proportionally more than testosterone did, the testosterone-to-cortisol ratio, a marker researchers use as a rough proxy for the body’s net anabolic state, actually declined by 14% at the highest dose.

That is the same two-numbers-at-once shape as the Finnish cohort data above, just compressed from years into an hour and swapping SHBG for cortisol as the counterbalancing hormone. In both datasets, the headline testosterone number moves in a direction that sounds favorable, while a second, related hormone shifts enough to work against it. Beaven’s team put the practical implication directly: caffeine’s testosterone boost “might be counteracted by the opposing catabolic effects of the increase in cortisol and resultant decline in the testosterone:cortisol ratio.”

What this evidence does not show

Neither study found here proves that coffee changes anyone’s usable testosterone in a way they would notice. The NFBC1966 analysis is cross-sectional and observational, meaning it measured coffee intake and hormone levels at a single point in time rather than tracking cause and effect, and the researchers themselves were explicit about the limits of that design: the findings “represent population-level” associations, and “reverse causation and residual confounding cannot be excluded.” Someone with naturally higher SHBG for unrelated reasons could simply be more likely to drink coffee, or a third factor, like diet or liver function, both of which influence SHBG independently, could be driving part of the association in both directions. The Oulu researchers’ own framing was similarly measured: the results “demonstrate associations rather than cause-and-effect relationships,” and they called for further research before the findings inform any dietary recommendation.

The Beaven exercise trial has a different, narrower limitation: it tested 24 professional rugby-league players, a small, young, highly trained, male-only sample, at caffeine doses, 800mg in particular, far above typical daily intake and above the roughly 400mg ceiling Mayo Clinic cites as the FDA’s general safety guidance for most healthy adults. It is also worth being precise about what “cup” means in the Finnish cohort’s own numbers: the study measured self-reported coffee servings, not milligrams of caffeine directly, so the per-cup changes in the table above should not be read as a precise per-milligram dose-response curve the way the controlled Beaven trial’s doses can be.

Worked example: weighing this against your own situation

1. Separate the two questions this page answers, the same way this site’s coffee and diabetes and coffee and blood pressure pages separate chronic from acute effects. “Does my daily coffee habit change my long-term SHBG and free testosterone” and “does a big pre-workout dose shift my testosterone-to-cortisol balance for a few hours” are different questions with different evidence behind them.
2. Do not treat a rising total testosterone number as automatically good news. The Finnish cohort data shows total testosterone and free testosterone can move in opposite directions in the same person at the same time, so a higher total reading alongside heavy coffee intake is not, on its own, evidence of more usable androgen.
3. If you already get bloodwork done, ask which testosterone measure is being reported. Per Mayo Clinic Laboratories’ own guidance, free or bioavailable testosterone is the more informative test specifically in situations, like unusually high or low SHBG, where total testosterone alone can be misleading.
4. Keep a pre-workout caffeine dose well under 800mg if testosterone-to-cortisol balance is a specific concern, since that was the dose where Beaven’s trial found cortisol’s proportionally larger rise actually pulled the ratio down, not up.
5. Do not expect to feel either effect. Nothing in either study measured a subjective symptom, muscle gain, libido change, or anything a person would notice day to day; both are hormone-panel findings, not performance or wellbeing outcomes.
6. If cutting back feels worth testing anyway, swap rather than quit. A decaf coffee selection from 1st in Coffee or the same category on Amazon keeps the ritual while removing most of the caffeine variable; this site’s own decaf coffee caffeine content breakdown has the real per-cup range, 2 to 15mg, for planning around it.

Common mistakes

Reading “coffee raises testosterone” as unambiguous good news is the most common mistake, and the full hormone panel in the Finnish cohort data directly contradicts a simple reading: total testosterone rose with coffee intake in men, while free testosterone and the free androgen index, the measures closer to what the body can actually use, both fell. Treating the cohort study as proof that coffee causes any of these hormone changes is the second mistake; the researchers’ own language is “associations,” not causation, and they explicitly flagged that reverse causation and other confounding factors cannot be ruled out in a cross-sectional design. Assuming the acute exercise findings apply to a normal morning cup is the third mistake; Beaven’s trial tested doses up to 800mg, roughly 8 cups of filter coffee, well above typical daily intake and the general 400mg safety ceiling. Assuming women show the same pattern as men is the fourth mistake; women in the cohort showed the SHBG-up, free-androgens-down half of the pattern with no significant change in total testosterone at all, a narrower result than men’s. Confusing a hormone-panel association with a symptom or performance outcome is the fifth mistake; neither study measured anything a person would actually feel or notice.

Common questions

Does coffee raise or lower testosterone? Both, depending on which measure is read. A 2026 study of 2,264 Finnish adults found higher coffee intake associated with higher total testosterone in men, but also higher SHBG, which in turn tracked with lower free testosterone and a lower free androgen index, the measures closer to usable testosterone.

What is SHBG and why does it matter here? Sex hormone-binding globulin is the main protein that transports testosterone in the blood. Per Mayo Clinic Laboratories, testosterone bound to SHBG is not readily available for tissues to use, so a rise in SHBG can push total testosterone measurements up while pushing free testosterone down in the same person, which is the exact pattern the coffee study found.

Does coffee affect women’s testosterone the same way it affects men’s? Not quite. Women in the Finnish cohort, excluding those with PCOS, showed a bigger per-cup rise in SHBG than men and a drop in free androgen measures, but no significant change in total or bioavailable testosterone at all, a narrower pattern than the one seen in men.

Does caffeine before a workout raise or lower testosterone? In the short term it raises it, but not cleanly. A 2008 trial in professional rugby players found an 800mg caffeine dose increased testosterone by a further 21% on top of exercise’s own effect, but also raised cortisol by 52%, enough that the testosterone-to-cortisol ratio, a rough marker of net anabolic state, actually declined.

Should I stop drinking coffee if I am worried about testosterone? Nothing in either study supports that conclusion. Both are observational or acute lab findings, not evidence of a meaningful real-world effect on symptoms, muscle mass, or libido, and the Finnish researchers themselves said more research is needed before the results should inform any dietary recommendation.

Is this the same mechanism as coffee’s effect on cortisol? Related but distinct. This site’s coffee and cortisol breakdown covers caffeine’s effect on cortisol directly, through the HPA axis. The acute testosterone study above shows cortisol and testosterone rising together after a large caffeine dose, with cortisol’s bigger rise working against testosterone’s gain in the net ratio, a related but separate hormonal interaction from the cortisol-only findings covered on that page.

For more sourced breakdowns like this one, delivered weekly, subscribe to The Sunday Pour.

Written by Theo Marsh · Published 10 October 2026