Peptides vs SARMs vs Steroids: Three Legal Categories, No Comparative Safety Data

Anabolic steroids are Schedule III controlled substances. SARMs are unapproved drugs FDA says cannot be legally marketed. Performance peptides are mostly unapproved too, and four sit on FDA's active safety-risk list. No study has compared the three classes.

The trade-off this page resolves is legal exposure against evidence quality against product integrity — three axes that point in three different directions, with no comparative literature to collapse them into a ranking. Most of the peptides discussed sit in the growth hormone peptides category, and the compound-level records are in tesamorelin: the only GHRH analogue still approved, ipamorelin: the selective secretagogue that failed its only published human trial and GHRP-2: the only growth hormone secretagogue that is an approved drug. This is a comparison, not a protocol: it carries no dosing, cycling or administration guidance for any of the three classes.

What actually separates peptides, SARMs and anabolic steroids?

Anabolic-androgenic steroids, SARMs and performance peptides are three receptor systems, not three grades of one drug. US statute defines an anabolic steroid as a substance "chemically and pharmacologically related to testosterone"; SARMs are non-steroidal ligands at that same androgen receptor; the peptides act on GHS-R1a, the GHRH receptor, IGF-1 receptors and the activin/myostatin axis. That divergence is why the three carry different legal categories, different evidence bases and different failure modes.

The classes get compared because they get sold to the same buyer, not because they are pharmacologically adjacent. A ghrelin-receptor agonist that provokes a growth hormone pulse and a non-steroidal androgen receptor ligand are doing unrelated things to unrelated tissues, and the only thing they reliably share is a marketplace.

That marketplace overlap is worth naming precisely, because one compound sits on both sides of it. Ibutamoren (MK-677) appears in the JAMA analysis below as a contaminant found inside products sold as SARMs, and it also sits on FDA's active list of bulk substances that may present significant safety risks alongside the four peptides named later on this page. A buyer who thinks they are choosing between categories may be holding a vial that ignores the distinction entirely.

Which of the three is a controlled substance, and what does that change?

Only anabolic steroids. DEA's controlled substances list, revised 25 June 2026, records "Anabolic steroids | 4000 | III" — Schedule III, with testosterone, nandrolone, oxandrolone, stanozolol and oxymetholone all under the same code. SARMs and the performance peptides are unapproved drugs, which is a regulatory category rather than a criminal one. That single distinction separates the three classes more sharply than any pharmacological difference between them.

The statutory definition is at 21 U.S.C. §802(41)(A): "The term 'anabolic steroid' means any drug or hormonal substance, chemically and pharmacologically related to testosterone (other than estrogens, progestins, corticosteroids, and dehydroepiandrosterone)…" That is the whole basis of the scheduling, and it is a chemical-relationship test rather than an effect test — which is precisely why non-steroidal SARMs fall outside it. The commonly cited name and year of the enabling statute could not be verified from a primary source in our record, so we cite the code section and the DEA list instead.

SARMs are barred from the market without being scheduled. FDA's consumer update states, verbatim: "SARMs cannot be legally marketed in the U.S. as a dietary supplement or drug at this time." Its fraudulent-products page, updated 2 December 2025, adds: "These products are unapproved drugs that FDA has not reviewed for safety and effectiveness." No named FDA official is quoted in either document, and the consumer update page carries contradictory dates, so we are not printing one. LiverTox states of the SARM class that "none were approved for human use", and USADA states that "all SARMs are for investigational purposes only". Approval outside the United States is not verified in our record, so every approval statement on this page is US-specific.

Anabolic steroids also have the most legitimate medical use of the three, which cuts against the usual framing. The Depo-Testosterone label reads: testosterone cypionate is "indicated for replacement therapy in the male in conditions associated with symptoms of deficiency or absence of endogenous testosterone", and the same label notes the product "contains testosterone, a Schedule III controlled substance". One caveat travels with that record: it was a repackager ANDA showing no boxed warning, and other testosterone products do carry boxed warnings — do not generalise from the one label to the class.

On the peptide side, tesamorelin is the class's only currently marketed FDA approval, at BLA 022505, and its scope is narrow: reduction of excess abdominal fat in HIV-infected adult patients with lipodystrophy. Everything else in the growth hormone group is either unapproved or approved somewhere else for something else. Jurisdictional detail is in are peptides legal and the movement of the FDA lists is tracked in the FDA peptide regulation timeline.

Which class has the strongest human evidence?

No SARM has ever passed a phase 3 trial, and none is in phase 3 today. Enobosarm, the best-studied member, ran POWER 1 (n=321) and POWER 2 (n=330) in muscle wasting, and its sponsor announced on 19 August 2013 that the trials "failed to meet the overall criteria for the co-primary responder endpoints". The peptide class holds one approved GHRH analogue with a 412-patient randomised trial behind it, and several compounds that failed the same way enobosarm did.

The POWER trials are the load-bearing fact for anyone told SARMs are "almost approved". Both completed in June 2014, both in muscle wasting in non-small-cell lung cancer, and the co-primary endpoints were lean body mass and stair-climb power. The day-84 responder analyses split: POWER 1 lean body mass p = 0.036 and stair-climb power p = 0.315; POWER 2 lean body mass p = 0.113 and stair-climb power p = 0.289. Three of the four missed. A peer-reviewed publication of the POWER 1 and POWER 2 results is not verified in our record, which is itself a fact about the evidence base — the design paper is published at PMID 27138015, the results are not.

What replaced them is smaller, not larger. Veru's current enobosarm programme is phase 2b, not phase 3: QUALITY has completed at 168 participants and PLATEAU is recruiting toward roughly 200, with an interim readout expected in the first quarter of 2027. Three enobosarm phase 3 breast-cancer trials were terminated, two of them after enrolling 52 and 5 patients. Twelve years after the phase 3 failure, the best-studied SARM is back in phase 2.

The peptide class is not exempt from this pattern, and the parallel is close. Ipamorelin reached phase 2 twice for postoperative ileus. The published trial missed its primary endpoint at p = 0.15 in 117 bowel-resection patients, and the larger 320-patient study completed in May 2014 and has never reported results. Follistatin is a 344-residue glycoprotein delivered as a gene-therapy payload rather than a peptide in a vial, and IGF-1 LR3, IGF-1 DES and PEG-MGF have zero registered human studies between them. The failure rate is comparable; only tesamorelin's approval separates the classes on this axis.

Are all three banned in sport, and where exactly?

All three are prohibited at all times, in and out of competition, and they sit in different sections of the 2026 list. WADA places anabolic androgenic steroids at S1.1 and SARMs at S1.2, both under S1 Anabolic Agents. The performance peptides are distributed across S2.2.4, S2.3, S4.3 and S2.2.1 according to mechanism. There are no lettered S1.1a or S1.1b subsections in the 2026 list — S1.1 is a single unified alphabetical list, and the exogenous/endogenous split is gone.

The S1.2 entry is worth reading verbatim, because it removes the most common defence offered for SARMs in tested sport: "Including, but not limited to: Clenbuterol, osilodrostat, ractopamine, selective androgen receptor modulators [SARMs, e.g. andarine, enobosarm (ostarine), LGD-4033 (ligandrol), RAD140, S-23 and YK-11], zeranol and zilpaterol." Six SARMs by name, in a section that applies at all times, in and out of competition.

The structural correction matters as much as the content. Older editions of the Prohibited List divided S1.1 into lettered subsections separating exogenous from endogenous anabolic androgenic steroids. The 2026 list does not, and a great deal of published material — and a great deal of automatically generated material — still reproduces the old lettering. Verified against two independent copies of the 2026 list effective 1 January 2026, the structure is S1.1 for AAS and S1.2 for other anabolic agents, with no letters beneath either. If a source cites S1.1a or S1.1b, it is citing an edition that no longer exists.

Peptides are not the exception people assume, either. Follistatin sits in S4, not S2, which is why athletes checking only the peptide-hormone section miss it. Detection is a separate question from prohibition, covered in do peptides show up on drug tests.

Which class has the worst product-quality record?

SARMs, by a wide margin, and the evidence is a single well-designed analysis rather than an impression. Van Wagoner and colleagues bought and analysed 44 products marketed as SARMs and found only 23 (52%) contained a SARM at all, 17 (39%) contained a different unapproved drug, 11 (25%) contained substances not on the label, and 4 (9%) contained no active compound whatsoever. The labelled amount matched the analysis in 18 of 44 products (41%).

The study is Van Wagoner RM, Eichner A, Bhasin S, and colleagues, "Chemical Composition and Labeling of Substances Marketed as Selective Androgen Receptor Modulators and Sold via the Internet," JAMA 2017;318(20):2004–2010 (PMID 29183075). The 44 products came from a survey of 51 supplier sites offering 210 products. Its conclusion, verbatim: "most contained drugs and unapproved substances. Only 52% contained selective androgen receptor modulators and many were inaccurately labeled."

The peptide market's own record is better on the measure we can test, and that is a narrower claim than it sounds. The Purity Index records a platform composite of 99.50 across 10,110 recency-weighted tests (verified August 2026), with ipamorelin at 99.65 across 323 recency-weighted tests from 21 laboratories (verified August 2026). Ipamorelin price data shows 63 shops in stock, a median of $6.00/mg and a low of $2.00/mg on a 10 mg vial (verified August 2026) — a 7× spread at nominally comparable purity. Those figures exclude shipping, taxes and customs, coupon codes, bulk tiers, multi-vial kits and account-gated pricing. Peptigrity tracks 530 shops and 11,852 independent lab tests across 118 peptides, with 1,283 community reviews (verified August 2026), and trust scores weight community reviews and independently verified HPLC purity equally at 50% each.

The two datasets answer different questions, and reading them as a scoreboard is a mistake. The JAMA analysis asked whether the labelled compound was present at all. HPLC purity asks what proportion of the material present is the intended peak. A vial can score 99% and hold the wrong quantity, the wrong salt form or the wrong species entirely — which is exactly what the peptide market's own failures look like. Follistatin has been tested seven times, and one May 2026 sample labelled 1 mg tested at 0.6 mg — a 40% shortfall. IGF-1 LR3 averages 98.51% across 125 tests, the lowest of any well-tested compound on the platform, with quantity variance from +1% to +72% and one June 2026 sample testing 72% over label on a compound whose signature adverse effect is hypoglycaemia. One sermorelin vendor's product measured 12.88 mg against a 10 mg label, a 28.8% overage.

The last row is the one that reframes the comparison. Independent batch-level testing exists for research peptides because a platform was built for it; nothing equivalent covers SARMs or anabolic steroids sold outside a pharmacy, which means the JAMA figures from 44 products are close to the whole of what is known. Individual peptide results are searchable in the lab test database, the method is in how to read peptide lab test results and mass spectrometry for peptides, and the fabrication patterns are in red flags in peptide certificates of analysis and why 10 mg isn't 10 mg.

What harms are documented for each class?

Each class has its own documented harm signature, measured in its own literature and never against the others. SARMs carry a LiverTox likelihood score of B — "likely cause of clinically apparent liver injury with jaundice". Anabolic steroid users showed LVEF 52±11% against 63±8% in 86 users versus 54 non-users. And four peptides — GHRP-2, GHRP-6, ipamorelin acetate and kisspeptin-10 — sit on FDA's active list of bulk substances that may present significant safety risks.

SARMs: the liver, the lipids and the HPG axis. LiverTox assigns the class a likelihood score of B, with latency "typically 2 to 3 months but ranged for a few weeks to a year". FDA's fraudulent-products page states, verbatim: "Life-threatening reactions, including liver injuries that required hospitalization, have occurred in people taking products containing SARMs." Its consumer update names heart attack and stroke, liver injury and acute liver failure, psychosis and hallucinations, sleep disturbance, sexual dysfunction, infertility, miscarriage and testicular shrinkage. Verified case reports include Mohamed WT and colleagues in Cureus 2023 (PMID 36945289) and Leung K and colleagues, "RAD-140 Drug-Induced Liver Injury", Ochsner Journal 2022 — a 24-year-old man who took RAD-140 for five weeks, reaching a peak bilirubin of 38.5 mg/dL, with biopsy showing "bile accumulation inside the hepatocytes and canaliculi", resolving after discontinuation. The LiverTox bibliography PMIDs were not individually verified in our record, so we are not publishing them.

The controlled data point the same way. Dalton JT, Barnette KG, Bohl CE and colleagues (Journal of Cachexia, Sarcopenia and Muscle 2011;2:153–161) gave enobosarm to 120 healthy elderly participants and measured lean body mass +1.3 kg over placebo at the top dose (p<0.001) alongside two costs: HDL fell dose-dependently, with "17% and 27% decreases noted at the 1- and 3-mg doses", and total testosterone fell in men by 6.4 nmol/L and 7.4 nmol/L (both p<0.001), with SHBG down and free testosterone, LH and FSH not significantly changed. LDL was not significantly affected. That is a low-dose, short, supervised study, and it is the floor of what unsupervised use does rather than the ceiling. Those figures are published trial exposures reported for their harm findings; they are not a protocol.

Anabolic steroids: the heart, and a liver distinction that is usually stated wrong. Baggish AL, Weiner RB, Kanayama G and colleagues, "Cardiovascular Toxicity of Illicit Anabolic-Androgenic Steroid Use," Circulation 2017 (PMID 28533317), compared 86 users against 54 non-users cross-sectionally: left ventricular ejection fraction 52±11% versus 63±8% (P<0.001), falling to 49±10% versus 58±10% among current users (P<0.001); early relaxation velocity 9.3±2.4 versus 11.1±2.0 cm/s (P<0.001); and coronary plaque volume at a median of 3 versus 0 mL³ (P=0.012), with lifetime dose associated with plaque burden at 0.60 SD per 10 years of use (P=0.008). The authors' conclusion: "Long-term AAS use appears to be associated with myocardial dysfunction and accelerated coronary atherosclerosis."

The hepatotoxicity claim needs its structural qualifier or it is simply inaccurate. LiverTox states that "the C-17α alkylated androgenic steroids have all been implicated in cases of liver injury" while "the C-17β esterified testosterones have only rarely been implicated in causing cholestasis". In plain terms: the oral 17α-alkylated compounds are the hepatotoxic ones, not the injectable esters. Four injury patterns are described — transient enzyme elevations, bland cholestasis, peliosis hepatis, and hepatic adenoma or hepatocellular carcinoma, typically after 5 to 15 years of use — with acute cholestasis occurring in roughly 1% of patients. A blanket "steroids damage the liver" collapses a structural distinction that determines whether the risk applies at all. On suppression and fertility, Vilar Neto JO and colleagues published a systematic review in Andrologia 2021 (PMID 33887077) asking whether AAS-induced hypogonadism is reversible; the citation is verified and its numeric findings are not, so no figures from it appear here.

Peptides: a regulator's active risk list, and it is not a short one. FDA's compounding bulk substances page, current 22 April 2026, lists GHRP-2, GHRP-6, ipamorelin acetate and kisspeptin-10 under "Bulk drug substances under category 2 of the interim policies" — substances that "may present significant safety risks", citing immunogenicity among other concerns. The agency's rationale for GHRP-2 is the sharpest single safety statement in the peptide record: "Compounded drugs containing GHRP-2 for injectable and nasal administration may pose risk for immunogenicity due to the potential for aggregation and peptide-related impurities. GHRP-2 also contains an unnatural amino acid, which adds to the complexity of peptide characterization. FDA is aware of reports of serious adverse advents in patients who received GHRP-2, including increased insulin requirement to maintain the blood glucose level, death of critically ill study subjects, infection and pancreatitis, though causality has not been established." The typographical error is FDA's and is preserved. For GHRP-6 the agency names "potential effect on cortisol and increase in blood glucose due to decreases in insulin sensitivity", and for kisspeptin-10 "immunogenicity for certain routes of administration" and "complexities with regard to peptide-related impurities and API characterization".

Cross-compound adverse-event context for the peptide side sits in the peptide side effects hub.

Has anyone actually compared the three classes?

No. No head-to-head or comparative safety literature exists across peptides, SARMs and anabolic-androgenic steroids. Searches of the record returned commercial content only. Every ranking of the three classes in circulation — including the ones that place peptides at the safe end and the ones that place steroids there — is assembled from three separate literatures with different populations, endpoints, durations and levels of medical supervision. There is no study to cite, in any direction.

The absence is structural rather than accidental, and understanding why matters more than lamenting it. The anabolic steroid evidence on harm is largely cross-sectional imaging in unsupervised long-term users — Baggish enrolled people already using, and measured them once. The SARM evidence is short, supervised, low-dose trials plus spontaneous case reports — Dalton dosed for a defined period under observation, and the liver cases arrived through publication rather than surveillance. The peptide evidence is small trials, several of them failed, plus a regulator's risk rationales. Those are three different measurement instruments applied to three different populations. Placing their outputs in one table produces a table, not a comparison.

This is where the honest answer refuses the question. A reader who wants to know which class is safest is asking something no published work has measured. The defensible statements are narrower and each one is class-specific: SARMs have the worst documented product-quality record; anabolic steroids have the only measured cardiac and coronary findings and the only criminal exposure; four peptides sit on a regulator's active safety-risk list and one peptide has an FDA approval. None of those facts ranks the classes against each other, because none of them was generated by a study that compared them.

So which should you choose?

There is no winner here, and the honest verdict is per-axis rather than overall. Anabolic steroids carry the only criminal exposure and the only measured coronary findings, and also the only routine legitimate medical use. SARMs have the worst product-quality record — 52% containing the labelled compound — and no successful phase 3 trial in over a decade. Peptides hold the largest independent testing dataset and four members on FDA's active safety-risk list.

The trial that would settle this

The study this comparison needs has never been run and is unlikely to be. It would require randomising healthy participants to compounds from three classes, two of which have no approved human indication and one of which is a controlled substance, with hard safety endpoints over years. No ethics committee approves that design. What exists instead are three separate literatures, and the specific gaps in each are listable even when the comparison is not.

Frequently Asked Questions

Are SARMs legal?

Not for sale as a supplement or a drug in the United States. FDA's position is that "SARMs cannot be legally marketed in the U.S. as a dietary supplement or drug at this time", and its fraudulent-products page describes them as unapproved drugs it has not reviewed for safety and effectiveness. They are not controlled substances, which is a different question from being legal to sell. Approval outside the US is not verified in our record.

Are peptides safer than SARMs or steroids?

Nobody knows, because no study has compared them. That is the accurate answer, and it is not a hedge. Four peptides — GHRP-2, GHRP-6, ipamorelin acetate and kisspeptin-10 — sit on FDA's active list of bulk substances that may present significant safety risks, and FDA's own GHRP-2 rationale cites deaths among critically ill study subjects and reports of pancreatitis, with causality unestablished.

Do anabolic steroids damage the liver?

Some do and some barely do, and the distinction is structural. LiverTox states that "the C-17α alkylated androgenic steroids have all been implicated in cases of liver injury" while "the C-17β esterified testosterones have only rarely been implicated in causing cholestasis". The oral 17α-alkylated compounds carry the hepatotoxicity; injectable esters largely do not. Injury patterns range from transient enzyme elevations to peliosis hepatis and, after 5 to 15 years, hepatic tumours.

Is any SARM close to FDA approval?

No. Enobosarm, the best-studied SARM, failed the co-primary endpoints of both its phase 3 trials in 2013, and its sponsor said so in writing. Three later phase 3 breast-cancer trials were terminated, two after enrolling 52 and 5 patients. The current programme is phase 2b, with an interim readout expected in the first quarter of 2027. No SARM is in phase 3 today.

Are all three banned in sport?

Yes, all of them, at all times. WADA's 2026 Prohibited List places anabolic androgenic steroids at S1.1 and SARMs at S1.2, naming andarine, enobosarm, LGD-4033, RAD140, S-23 and YK-11 explicitly. The performance peptides sit at S2.2.4, S2.3, S4.3 and S2.2.1 depending on mechanism. The lettered S1.1a and S1.1b subsections that older sources cite do not exist in the 2026 list.

What are the odds a SARM bought online is what it claims?

Roughly even, on the only systematic analysis available. Of 44 products bought from online suppliers and analysed for JAMA, 23 (52%) contained a SARM, 17 (39%) contained a different unapproved drug, 11 (25%) contained substances not on the label, and 4 (9%) contained no active compound at all. The labelled amount matched the analysis in 41% of products.

Where this leaves the comparison

The three classes do not sort into a safety ranking, and the strongest thing this page can say is that no study has ever tried to produce one. What does sort cleanly is legal category: anabolic steroids are Schedule III controlled substances under 21 U.S.C. §802(41)(A), SARMs are unapproved drugs FDA says cannot be legally marketed, and the performance peptides are unapproved drugs of which four sit on a regulator's active safety-risk list. Those are three different kinds of exposure, and only one of them is criminal.

On product integrity the answer is unusually clear. Only 52% of 44 SARM products contained the compound on the label and 9% contained nothing at all, against a peptide market with 11,852 independent lab tests and a composite purity index of 99.50. That comparison is real, and it is narrower than it looks: HPLC purity and "does this contain the labelled drug" are different measurements, and the peptide market's own failures — a 40% shortfall on follistatin, a 72% overage on IGF-1 LR3 — are quantity failures that a purity number never shows.

None of this makes peptides the safe choice, and this page should not be read that way. The best-studied SARM failed its phase 3 endpoints and ipamorelin failed its phase 2 endpoint; four peptides carry a regulator's significant-safety-risk designation and FDA's own file on GHRP-2 records deaths among critically ill study subjects with causality unestablished; and all three classes are prohibited in sport at all times. The comparison that would settle any of this does not exist, and saying so is more useful than manufacturing a ranking out of three unrelated literatures.

Browse the growth hormone peptides category, or our complete peptide guide with 118 compounds (verified August 2026). For per-injection volume, use the peptide dosing calculator alongside the reconstitution calculator. Compare shops through independent lab tests and community-verified shop reviews.

This article is for educational and informational purposes only and does not constitute medical advice. Peptides discussed may be investigational compounds not approved by the FDA (or equivalent regulators in your jurisdiction) for human use. Always consult a qualified healthcare provider before using any peptide or research compound. Peptigrity is an independent review platform and does not sell, endorse, or recommend specific products or vendors.