Testosterone therapy treats hypogonadism, or low testosterone, which is a deficiency of the primary male sex hormone that can cause symptoms such as low libido, fatigue, and erectile dysfunction. The American Urological Association (AUA) defines low testosterone as a total testosterone level consistently below 300 ng/dL on at least two early-morning blood tests, considered together with relevant symptoms or signs.1 Testosterone prescribing rose sharply in the United States through the 2000s and early 2010s, a period marked by direct-to-consumer “low T” marketing and the wider availability of convenient topical formulations.2 As use expanded, concerns about cardiovascular safety emerged, and in 2015 the U.S. Food and Drug Administration (FDA) required labeling that warned of a possible increased risk of heart attack and stroke.3 Prescribing subsequently declined.4 In 2023, the TRAVERSE trial found that, among men with hypogonadism and pre-existing elevated cardiac risk, testosterone therapy was not associated with an additional increase in the risk of major adverse cardiac events compared with placebo.5 In February 2025, the FDA removed boxed-warning language about an increased risk of adverse cardiovascular outcomes from testosterone product labels, while adding a class-wide warning about increased blood pressure.6 Prescribing has since rebounded to and beyond its early-2010s levels.7 As new prescriptions rise again, it is unclear whether testing to establish a deficiency before treatment has kept pace.
We studied 300,277 male patients who were ordered a new testosterone prescription between January 2017 and June 2026. For each patient starting testosterone, we determined whether the prescription was preceded, in the prior 12 months, by any testosterone test, by two or more tests on separate days, by at least one low total-testosterone result (below 300 ng/dL), or by two low total-testosterone results within six months beforehand (our proxy for the lab-observable AUA testing criterion).
Testing before starting testosterone became more common over the study period, but testing that meets the lab-observable portion of the guideline standard for diagnosing testosterone deficiency remained uncommon. The share of men starting testosterone who had at least one testosterone test in the prior 12 months rose from 72.3% in 2017 to a peak of 83.8% in 2024 before declining slightly to 82.6% in the first half of 2026. The proportion of men with a new testosterone prescription who had two or more tests on separate days in the year prior to the prescription rose fairly steadily, from 28.5% in 2017 to 40.8% in the first half of 2026. The share meeting the lab-observable AUA criterion (two total-testosterone results below 300 ng/dL on separate days) within six months before prescribing rose from 16.9% in 2017 to a peak of 22.8% in 2024 to 21.8% in the first half of 2026. New prescriptions are skewed toward more recent years, consistent with the rise in prescribing previously reported.7
The presence of a documented low testosterone result before starting testosterone was common across all age groups but declined modestly with age. Among men who started testosterone, the share with at least one total-testosterone result below 300 ng/dL in the prior year ranged from 66.2% among those aged 30–39 to 58.8% among those aged 75 and older.
Testing in the year prior to starting testosterone varied by prescribing specialty, and the specialties writing the most prescriptions were not the ones testing most consistently. We grouped prescribers into primary care, urology, endocrinology, and oncology/hematology, with all remaining specialties combined. Urology and endocrinology are specialties that often diagnose and manage testosterone deficiency; oncology and hematology encounter it in narrower contexts, since some cancer treatments lower testosterone.8 Most men starting testosterone received their prescription from primary care (57.6%), with urology (17.4%) and endocrinology (11.0%) a distant second and third and oncology/hematology accounting for fewer than 1%. Among these groups, prescriptions from endocrinology had the highest share with a low testosterone result in the prior year, at 69.6%, while urology had the lowest, at 60.0%.
Testosterone therapy is often advertised via direct-to-consumer channels, men who are prescribed testosterone this way are likely not included in this study.
