Selective androgen receptor modulator (SARM)

S-23

S-23 had a terminal half-life of 10.9 h (intravenous) and 11.9 h (oral) in male rats; the app lists 12 hours, and no human plasma half-life was identified.1

Emerging evidence (the app's grade, not a conclusion of the sources)Get the app

Level after one dose

S-23 after one dose: 50% at ~12 h, 25% at 24 h, 12.5% at 36 h.

Drawn from the half-life in the app's model, ~12 h. The model's figures

S-23 is a selective androgen receptor modulator (SARM) that Jones and colleagues at The Ohio State University and GTx Inc. described in male rats as an agent for hormonal male contraception, given with estradiol benzoate by subcutaneous injection in the contraception study (Endocrinology, 2009).1 The human studies identified in PubMed are urinary elimination studies done for doping control.2,3 Ameline and colleagues (Journal of Pharmaceutical and Biomedical Analysis, 2022) followed one volunteer after a single oral dose of approximately 8 mg, and Alhalabi and colleagues (Biomedical Chromatography, 2025) followed five healthy male volunteers per group after single and repeated microdoses of 1, 10 and 50 μg.2,3

The World Anti-Doping Agency's 2026 Prohibited List names S-23 as an example under S1.2, other anabolic agents.4 No label for S-23 was identified: a DailyMed query on 16 November 2026 returned 0 labels.3 The app grades the evidence for S-23 as emerging.

The measurements below are from rat studies and cell assays, not from people. In the Jones 2009 binding assay the inhibitory constants were 0.45 ± 0.2 nm for dihydrotestosterone and 1.7 ± 0.2 nm for S-23, and S-23 was identified as a full agonist at the androgen receptor in a cell assay.1 In 30 castrated male rats (n = 5 per group) given daily subcutaneous S-23 for 14 days, the ED50 was 0.43 ± 0.18 mg/d in the prostate, 0.41 ± 0.0070 mg/d in the seminal vesicles and 0.079 ± 0.010 mg/d in the levator ani muscle.1 The authors write that "at a dose rate as low as 0.1 mg/d, S-23 was able to selectively maintain the weight of the levator ani muscle at the intact control level, whereas its effects on the prostate and seminal vesicles were lower than 30% of those observed in intact controls".1 A later paper from the same group (Jones and colleagues, Journal of Pharmacology and Experimental Therapeutics, 2010) compared a series of SARMs in castrated male rats and wrote that S-23 "displayed full agonist activity in androgenic and anabolic tissues; however, the remaining SARMs were more prostate-sparing".5 In ovariectomized female rats in that paper, ovariectomy-induced rises in follicle-stimulating hormone (FSH) were inhibited by testosterone propionate and all the SARMs, "with the greatest reduction again observed with TP and S-23".5 For the class as a whole, Alhalabi and colleagues write that in vitro experiments "have shown that SARMs exhibit agonistic activity on the androgen receptor, primarily in muscle and bone tissues while demonstrating reduced activity within other tissues such as prostate and seminal vesicles".3

3

Half-life and pharmacokinetics

Values from the Milligram app's model.

Half-life
~12 h
Bioavailability
96%
Typical dose
10–30 mg/day
Route
Oral
Time to peak
~2 h

The second column of the table holds the app's own figures, shown beside what the sources report.

Table 3.1Half-life and pharmacokinetics5 rows
ParameterApp's figureReported in the sources
Half-life~12 hMale rats given 10 mg/kg: terminal half-life 10.9 h after an intravenous dose (n = 4) and 11.9 h after an oral gavage dose (n = 5), with no significant difference between the two; no human plasma half-life identified1
Bioavailability96%Oral bioavailability of 96% in male rats after 10 mg/kg by gavage, "indicating its near to complete absorption after oral dose"; none reported in people1
Time to peak~2 hMale rats given 10 mg/kg by gavage: time to peak 252 ± 137 min (about 4.2 h), with maximum plasma concentrations forming "a plateau over 6–10 h after dose"1
Clearance and volume of distributionNot set in the appMale rats: clearance 0.87 mL/min per kg and steady-state volume of distribution 655 mL/kg after the intravenous dose1
Urinary elimination in peopleNot set in the appParent S-23 detectable in urine for an average of up to 253 h after a single 1 μg dose and up to 544 h after a single 50 μg dose in five healthy male volunteers per group3

The app's model type is one-compartment, whereas Jones and colleagues analysed the rat plasma data "using noncompartmental methods".1 In the one volunteer studied by Ameline and colleagues, after a single oral dose of approximately 8 mg the parent drug "was detectable in hydrolyzed urine from 2 h post administration up to 28 days, with concentrations ranging between 0.5 and 93 ng/mL".2 Alhalabi and colleagues report that after a single 50 μg dose intact S-23 "was detected in urine for 544 h following a single oral 50 μg dose, with an average peak concentration of 1790 pg/mL reached after 18.5 h" in urine.3 These are urinary peak times, not plasma times to peak, and they are not comparable with the app's 2 hours.3

4

Human studies

No clinical trial of S-23 in people measuring efficacy or safety was identified in PubMed or ClinicalTrials.gov: a ClinicalTrials.gov query for "S-23" returned 14 records, none of them about this compound.

Table 4.1Human studies3 rows
Study and yearDesignPeopleDose and lengthMain result
Ameline and colleagues, J Pharm Biomed Anal 20222Controlled single oral administration with urine analysis; in vitro human liver microsome experiment alongside1 volunteerApproximately 8 mg once; urine followed up to 28 daysParent drug detectable in hydrolyzed urine from 2 h to 28 days at 0.5 to 93 ng/mL; hair sample one month later negative2
Alhalabi and colleagues, Biomed Chromatogr 2025, single dose3Single-dose urinary elimination study; no comparator group5 healthy male volunteers per dose group1, 10 or 50 μg orally once, in 50 mL drinking yoghurt; urine collected for 30 daysParent drug detectable for an average of up to 253 h (1 μg) and up to 544 h (50 μg)3
Alhalabi and colleagues, Biomed Chromatogr 2025, repeated dose3Repeated-dose urinary elimination study; no comparator group5 healthy male volunteers per dose group1, 10 or 50 μg orally on 5 consecutive days; urine collected for 30 days after the last doseEach peak after consecutive doses was higher than the previous; parent drug detected for an average of 568 h after 5 × 50 μg3

Deeper analysis: Alhalabi and colleagues, 2025

Alhalabi and colleagues at the Center for Preventive Doping Research, German Sport University Cologne, published the urinary elimination study in Biomedical Chromatography in 2025 to help separate inadvertent intake from intentional use.3 Design: S-23 dissolved in ethanol was added to 50 mL of drinking yoghurt and given by mouth, urine was analysed by liquid chromatography with tandem mass spectrometry after enzymatic hydrolysis, and the work was done "with ethical approval granted by the local ethics committee" and written informed consent.3 The measurements were urinary concentrations and detection windows, with a limit of detection of 1 pg/mL.3 The authors report that "A total of 18 metabolites was detected in human in vivo samples following oral administration of microdosed S-23".3 Results: the longest-detectable analyte was the unchanged drug, and after the single 50 μg dose its average urinary peak was 1790 pg/mL at 18.5 h, with 5%–10% of the dose excreted within 24 h across doses.3 The paper's Table 3 gives maximum detection windows as ranges: 213–360 h after a single 1 μg dose, 429–544 h after a single 50 μg dose and 500–568 h after 5 × 50 μg, and its results text gives 193 h for the maximum window after 1 μg, so the 253 h and 544 h figures are averages.3 Each excretion profile showed a later, smaller second peak, which the authors suggest "could be explained by a high protein binding behavior".3 The authors attribute the long detection times to the assay: "Noteworthy is that these prolonged detection times are attributed to the low limit of detection (LOD) of 1 pg/mL".3 Limitations the paper states or shows: "The potential influence of the yogurt vehicle on the bioavailability of" S-23 was not assessed, absolute quantification was not accomplished for the metabolites, there was no comparator group, and each dose level had five volunteers.3 The method was validated for semiquantitative detection, the paper reports distinct interindividual variability in excretion over the first 5 days of the multiple-dose study, and its elimination-profile figure is drawn from 5, 4 or 3 underlying data points.3 The study reports concentrations and detection times only; the text read contains no adverse-event outcomes.3 Funding and ties: the authors state "This work was supported by Bundesministerium des Innern, für Bau und Heimat and Manfred Donike Institut für Dopinganalytik" and "The authors declare no conflicts of interest."3

Deeper analysis: Ameline and colleagues, 2022

Ameline and colleagues at the Institut de Médecine Légale in Strasbourg reported the in vitro and single-volunteer work in the Journal of Pharmaceutical and Biomedical Analysis in 2022; only the PubMed record's summary text was read, not the full paper.2 In human liver microsomes four metabolites were produced, and in the volunteer's urine "only one of the four metabolites identified in vitro was detected, hydroxy-S-23".2 The authors also report that another metabolite, dihydroxy-S-23, not identified in vitro, was identified in the urine of the volunteer.2 The authors state that "nothing has been published about long-term effects of S-23".2 One volunteer cannot show how results vary between people, and the record gives no funding statement or author ties.2

Deeper analysis: Jones and colleagues, 2009 rat study (animals, not people)

Jones and colleagues reported the rat work in Endocrinology in 2009, with affiliations at The Ohio State University, the University of Tennessee and GTx Inc.1 Design: "Forty-two male Sprague Dawley rats (90 d old) were randomly assigned to seven groups (n = 6/group)", given daily subcutaneous injections for 70 days of vehicle, estradiol benzoate (EB) 5 μg/d alone, or EB plus S-23 at 0.05, 0.1, 0.3, 0.5 or 0.75 mg/d.1 Estradiol was included because male rats need it for normal sexual behaviour, and mating trials were done in only four of the seven groups.1 Main result: at 0.1 mg/d with EB, "four of six animals showed no sperm in the testis and zero pregnancies (none of six) in mating trials".1 Secondary and contrary results: the effect on sperm was biphasic, and "Higher doses of S-23 actually supported spermatogenesis in a dose-dependent manner, with the 0.75 mg dose resulting in a similar mean sperm count as those in the EB alone group and intact control group".1 In the mating trials, four of five rats given estradiol benzoate alone were fertile, and one of six rats given estradiol benzoate plus 0.3 mg of S-23 was fertile.1 Serum luteinizing hormone (LH) fell below the assay detection limit with drug treatment, FSH was significantly suppressed only at the higher doses, and testis weights were significantly lower in all S-23 groups.1 Fat mass fell with dose ("Coadministration of S-23 with EB decreased FM in a dose-dependent manner"), but the results text states that S-23 "did not result in dose-dependent effects on body weight or BMD", while the summary text says S-23 "increased bone mineral density and lean mass".1 Reversibility: "an 83% fertility rate was achieved 70 d after termination of treatment; however, infertility was fully reversible after 100 d, with a 100% pregnancy rate", and prostate weights "remained below that of control".1 Limitations in the authors' words: "Unfortunately, we were unable to develop analytical methods sufficient to measure intratesticular concentrations of S-23 to confirm this hypothesis", "Histological analysis of testis and epididymides from animals after drug treatment is ongoing in our laboratory", and the work shows "a very narrow therapeutic dose range to achieve no sperm in the testis and infertility".1 The paper closes that the SARM characteristics "represents an important step toward the male pill, but much remains to be done".1 Funding and ties: the work "was supported by a grant from National Institute of Diabetes and Digestive and Kidney Diseases Grant 1 R01 DK59800-05"; A.J. has stock options with GTx, Inc.; D.J.H., D.D.M. and J.T.D. are employed by and have stock options with GTx, Inc., are inventors on U.S. Patent 6492554 and have received patent royalties; and J.C. is an inventor on U.S. patent 6492554 and has received patent royalties.1

5

Doses reported in studies and on labels

No label for S-23 was identified, so the table lists doses from the studies read.

Table 5.1Doses reported in studies and on labels6 rows
SourceDose and routeLengthWho or what, and comparator
Ameline 2022 (people)2About 8 mg, single oral doseOnce; urine followed up to 28 days1 volunteer; no comparator
Alhalabi 2025, single dose (people)31, 10 or 50 μg, oral, in drinking yoghurtOnce; urine for 30 days5 healthy male volunteers per dose; no comparator
Alhalabi 2025, repeated dose (people)31, 10 or 50 μg, oral, in drinking yoghurtDaily for 5 days; urine for 30 days after the last dose5 healthy male volunteers per dose; no comparator
Jones 2009, pharmacokinetics (male rats)110 mg/kg, intravenous or oral gavageSingle dose; samples to 48 hn = 4 intravenous, n = 5 oral
Jones 2009, castrated rats (male rats)10.01, 0.05, 0.1, 0.5, 1 or 3 mg/d, subcutaneous14 daysn = 5 per group; castrated and intact vehicle controls
Jones 2009, contraception study (male rats)10.05, 0.1, 0.3, 0.5 or 0.75 mg/d subcutaneous with estradiol benzoate 5 μg/d70 daysn = 6 per group; vehicle and estradiol-only controls

The app's typical dose is 10 to 30 mg/day by mouth; none of the studies read gives a daily dose in that range for people, the human studies used about 8 mg once or 1 to 50 μg, and the rat doses were in mg/kg or mg/d per animal.

6

Side effects reported

No trial of S-23 in people reporting adverse events was identified, so there are no human adverse-event rates to give. Ameline and colleagues state that "nothing has been published about long-term effects of S-23".2 In rats given estradiol benzoate with S-23, Jones and colleagues observed lower testis weights in every S-23 group, LH below the assay detection limit with drug treatment (estradiol benzoate alone also decreased LH), FSH significantly suppressed only at the higher doses, and prostate weights that "remained below that of control" after treatment stopped.1 For SARMs as a class, not for S-23 specifically, Alhalabi and colleagues write that "case reports and clinical studies have described several side effects" and that "The observed side effects comprise serum and lipid level abnormalities, potential increase in the risk of cardiovascular diseases and liver damages as well as potentially reduced testosterone levels".3

7

Bloodwork and markers

The app lists total testosterone, lipids and liver enzymes as markers for S-23. In the rat studies the hormones measured were serum LH and FSH and intratesticular testosterone, and the human studies measured urinary drug and metabolite concentrations.1,3 None of the S-23 papers read reports serum total testosterone, lipid or liver enzyme results for S-23 in people.1,3,2 The class-level statement quoted in the side-effects section links SARMs generally with lipid abnormalities, liver damage and reduced testosterone; none of the sources read discusses monitoring these markers for S-23.3

8

Regulatory and anti-doping status

The WADA 2026 Prohibited List names S-23 in S1.2, other anabolic agents, within S1, anabolic agents, which it lists as "PROHIBITED AT ALL TIMES (IN- AND OUT-OF-COMPETITION)".4 The list states that all prohibited substances in S1 "are non-Specified Substances".4 Alhalabi and colleagues write that "S-23 has been identified in five AAFs reported in 2022", where AAF is an adverse analytical finding in doping control, and that inadvertent exposure through contaminated dietary supplements "has emerged as a significant concern".3 A DailyMed query on 16 November 2026 returned no label for S-23.3

  • No human plasma half-life, bioavailability or time to peak for S-23 was identified in PubMed; the app's 12 hours and 96% match or sit near the rat figures (10.9 to 11.9 h, 96%), whereas the app's 2 hours differs from the rat time to peak of about 4.2 h.1
  • No controlled trial of efficacy or safety in people was identified in PubMed or ClinicalTrials.gov, so the app's description of S-23 as a male contraceptive under investigation rests on rat work in the sources read.1
  • Long-term effects in people are not reported.2
  • Whether the yoghurt vehicle changed absorption in the microdose studies was not assessed.3
  • The app's 10 to 30 mg/day has no matching dose in the sources read.
What is the half-life of S-23?

In male rats the terminal half-life of S-23 was 10.9 h after an intravenous dose and 11.9 h after an oral dose, and the app lists 12 hours.1 No human plasma half-life was identified in PubMed.

What is the bioavailability of S-23?

Oral bioavailability was 96% in male rats given 10 mg/kg by gavage, and the app lists 96%.1 No human figure was identified in PubMed.

How long does S-23 take to reach its peak level?

In male rats given 10 mg/kg by gavage the time to peak was 252 ± 137 min, about 4.2 hours, and the app lists about 2 hours.1 Jones and colleagues describe maximum plasma concentrations "forming a plateau over 6–10 h after dose".1

Is S-23 banned by WADA?

Yes, the WADA 2026 Prohibited List names S-23 among the SARMs under S1.2, other anabolic agents, prohibited at all times.4

Is S-23 approved?

No label or approval for S-23 was identified: a DailyMed query on 16 November 2026 returned 0 labels.3 Alhalabi and colleagues write that "no SARM has been approved or commercialized as a pharmaceutical drug".3

How long can S-23 be detected in urine?

In five healthy male volunteers per dose, a single 1 μg dose was detectable for an average of up to 253 h and a single 50 μg dose for up to 544 h.3 The authors attribute the long windows to a limit of detection of 1 pg/mL.3

Has S-23 been studied as a male contraceptive?

Only in rats among the sources read: with estradiol benzoate by daily subcutaneous injection, 0.1 mg/d of S-23 gave four of six animals with no sperm in the testis and none of six pregnancies in mating trials, whereas one of six rats given 0.3 mg/d was fertile and four of five given estradiol benzoate alone were fertile.1 Higher doses supported sperm production, and infertility was fully reversible after 100 days of recovery in the rats.1

Has S-23 been tested in people?

The human studies identified are urinary elimination studies, one volunteer given about 8 mg and five volunteers per group given 1 to 50 μg, and no trial of efficacy or safety was identified in PubMed or ClinicalTrials.gov.2,3

12

References

  1. Jones A, Chen J, Hwang DJ, Miller DD, Dalton JT. Preclinical characterization of a selective androgen receptor modulator for hormonal male contraception (S-23). Endocrinology 2009;150(1):385-395. PMID 18772237
  2. Ameline A, Gheddar L, Raul JS, Kintz P. In vitro characterization of S-23 metabolites produced by human liver microsomes, and subsequent application to urine after a controlled oral administration. J Pharm Biomed Anal 2022;212:114660. PMID 35182830
  3. Alhalabi H, Korsmeier L, Thomas A, Thevis M. Investigations into the urinary metabolite elimination profile of the selective androgen receptor modulator S-23 in studies mimicking contaminated product ingestion for doping control purposes. Biomed Chromatogr 2025. PMID 40277337
  4. World Anti-Doping Agency. World Anti-Doping Code International Standard: Prohibited List 2026. WADA list
  5. Jones A, Hwang DJ, Duke CB 3rd, He Y, Siddam A, Miller DD, Dalton JT. Nonsteroidal selective androgen receptor modulators enhance female sexual motivation. J Pharmacol Exp Ther 2010. PMID 20444881