MuscleLedger

Peptides vs Steroids: The Comparison Nobody Has Actually Run

Anabolic steroids were tested in trained men against strength endpoints in 1996. No muscle peptide has ever been tested that way — which is the real finding, and it cuts both ways.

Leon H · Published 2026-09-08

This is the question the market is really asking, and it is usually answered by whoever is selling. Both answers on offer are wrong: the vendor claim that peptides do what steroids do without the downside, and the dismissive claim that peptides do nothing at all. The evidence supports neither, because the evidence is lopsided in a specific way that is worth understanding.

This is research journalism. It contains no usage guidance, this site publishes no dose figures, and dosing questions are answered by Peptifact rather than here.

One class was tested properly. The other was not.

In July 1996 the New England Journal of Medicine published the trial that settled the steroid question. Forty-three normal men were randomly assigned to four groups: placebo without exercise, testosterone without exercise, placebo plus exercise, and testosterone plus exercise. Supraphysiologic weekly testosterone injections or placebo for ten weeks; the exercise groups performed standardised weight-lifting three times weekly. Fat-free mass was measured by underwater weighing, muscle size by MRI, and strength by bench press and squat.

Group Fat-free mass Bench press Squat
Testosterone + exercise +6.1 ± 0.6 kg +22 ± 2 kg +38 ± 4 kg
Testosterone, no exercise Triceps area +424 ± 104 mm² +9 ± 4 kg +16 ± 4 kg
Placebo, no exercise Triceps area −81 ± 109 mm² −1 ± 1 kg +3 ± 1 kg

Two findings, both robust. Testosterone plus training produced a large effect on every endpoint. And testosterone without any training at all beat placebo without training on muscle size and on both strength measures — the drug did something on its own, in men who did not lift.

That is what a properly designed muscle trial looks like: the right population, the right endpoint, randomised, placebo-controlled, blinded, with the training variable factorially separated.

No peptide sold for muscle growth has ever been tested this way. Not in trained adults. Not against a one-repetition-maximum endpoint. Not once, in any country, in thirty years.

What the peptide trials measured instead

Our compound pages walk through each record individually. In aggregate:

Compound Human trials with a muscle endpoint Population Result
MK-677 1 (12 months) Adults 60–81, not training +1.1 kg fat-free mass; no strength or function change
Capromorelin 1 (12 months) Adults 65–84 +1.4 kg lean mass; +0.9 s tandem walk; terminated early
CJC-1295 0 Only trial measured hormone concentrations
Ipamorelin 0 Only trials were in bowel-surgery patients
SS-31 / elamipretide 1 (7 days) Elderly, energetics endpoint No separation from placebo
BPC-157, TB-500 0 No completed human trial of either — see the Wolverine stack

The instinct is to line up 6.1 kg against 1.1 kg and declare the contest over. Do not do that. The testosterone figure comes from young men who were lifting weights three times a week; the MK-677 figure comes from adults aged 60 to 81 who were not. Those are different questions asked of different people, and the arithmetic comparison is meaningless.

The meaningful comparison is structural, and it is worse for the peptides than the arithmetic one: for one class the decisive trial exists, and for the other it has never been attempted.

Why nobody ran it

This is not an oversight, and understanding why is the useful part.

Testosterone was studied because it is an approved drug with a manufacturer, a patent history and clinical indications, and because a research question about androgens and muscle had funding attached to it. The 1996 trial was run to answer whether athletes' claims were substantiated — the paper opens by noting the efficacy of androgenic steroids for strength was "unsubstantiated" at the time.

The muscle peptides have no such sponsor. Ipamorelin's two trials were run by a pharmaceutical company for a gastrointestinal indication and stopped when that indication failed. CJC-1295's sole registration was terminated in a fat-distribution study and never reported. Both compounds passed out of pharmaceutical development and into a direct-to-consumer market where nobody has an incentive to fund a trial that could produce a negative answer — and where the product sells regardless.

The two exceptions prove the rule: MK-677 and capromorelin were both developed as oral drugs for age-related muscle loss by companies that needed a regulator's approval, so both were tested against real endpoints in adequately sized trials. Those are the two compounds in this whole market with credible functional data, and both produced modest results.

Safety: an asymmetry that runs the other way

Steroids are the class with the documented harms, and it is a long list: cardiovascular, hepatic, endocrine suppression, psychiatric effects, all built up over decades of study and use.

It does not follow that peptides are safer. It follows that they are less studied. A compound with two randomised trials cannot have an adverse-effect profile, in the same way a compound with no trials cannot have an efficacy profile — the absence in both directions has the same cause.

Where these compounds have been studied, a consistent signal appears and it is metabolic. In the twelve-month MK-677 trial, fasting glucose rose about 5 mg/dL, insulin sensitivity fell and cortisol rose; the capromorelin trial reported the same direction of travel with fatigue and insomnia added. A drug class that raises appetite while reducing insulin sensitivity is making a specific trade, and it appears to be a class property rather than a quirk of one study.

FDA reached its own conclusion about several of these compounds independently of any efficacy question: ipamorelin and ibutamoren both sit in category 2 of the agency's bulk drug substances lists, the tier reserved for substances presenting significant safety risks, which is why neither can be lawfully compounded.

Both classes are prohibited in tested sport, at all times, in and out of competition — anabolic agents under WADA section S1, peptide hormones, growth factors and related substances under S2. Neither is a loophole. For several muscle peptides the published detection literature is larger than the clinical literature, a pattern documented on our drug testing page and visible in the record of every compound page linked above.

The legal position differs in type. In the United States anabolic steroids are Schedule III controlled substances under the Anabolic Steroid Control Act. The muscle peptides are generally not scheduled; they are unapproved drugs, sold labelled for research use. That is a different category of legal exposure rather than an absence of one, and it deserves reading on its own terms rather than being treated as a point scored in a comparison.

What an honest reader should take away

The comparison in the title cannot be made from evidence, and saying so is more useful than faking an answer.

What can be said, with sources behind every clause: anabolic steroids have a decisive randomised trial in the exact population that buys them, with a large effect on mass and strength, and a long documented harm profile. The muscle peptides have two credible functional trials between them, both in older non-training adults, both producing about a kilogram of mostly fat-free mass, one with no strength change at all — plus a consistent metabolic signal, a category 2 safety listing on several compounds, and a ban in tested sport that applies year-round.

Anyone who tells you peptides are a milder route to the same place is describing a mechanism, not a result. The result has never been measured. Our evidence ledger grades every compound in this market on that basis, and the compound this site rates highest for evidence is not a peptide at all.

Sources

  • Bhasin S, Storer TW, Berman N, et al. The effects of supraphysiologic doses of testosterone on muscle size and strength in normal men. N Engl J Med 1996;335(1):1–7. PMID 8637535
  • Nass R et al. Effects of an oral ghrelin mimetic on body composition and clinical outcomes in healthy older adults. Ann Intern Med 2008. PMID 18981485
  • White HK et al. Effects of an oral growth hormone secretagogue in older adults. J Clin Endocrinol Metab 2009. PMID 19174493
  • Beck DE et al. Proof-of-concept study of the ghrelin mimetic ipamorelin. Int J Colorectal Dis 2014. PMID 25331030
  • Teichman SL et al. Prolonged stimulation of GH and IGF-I secretion by CJC-1295. J Clin Endocrinol Metab 2006. PMID 16352683
  • WADA Prohibited List, sections S1 and S2.
  • FDA, Bulk Drug Substances Nominated for Use in Compounding — category 2.
  • ClinicalTrials.gov searched 2026-09-08 for registered trials of muscle peptides in trained adults: none found.

Frequently asked questions

Are peptides as effective as steroids for building muscle?

There is no trial that answers this, and the reason is one-sided. Anabolic steroids were tested in exactly the population and design that would settle the question — young men lifting weights, randomised, placebo-controlled — and produced 6.1 kg of fat-free mass in ten weeks. No peptide sold for muscle has ever been tested in that population under any design. Comparing the numbers that do exist means comparing 22-year-old lifters to 70-year-old non-exercisers, which is not a comparison. What can be said is that the peptide result in the population where it was measured was about a kilogram over a year, with no strength change.

Are peptides safer than steroids?

Safer is not the same as safe, and the honest answer has two halves. Anabolic steroids have a large, well-characterised adverse-effect literature built over decades — cardiovascular, hepatic, endocrine, psychiatric — because they have been studied and used at scale for that long. Most muscle peptides have no comparable literature, and an absence of documented harms in a compound with two randomised trials is not a safety record. Where secretagogues have been studied, a consistent metabolic signal appears: raised appetite alongside reduced insulin sensitivity and higher fasting glucose.

Do peptides show up on a drug test the way steroids do?

Yes, and often through more recently developed methods. Anabolic agents sit in section S1 of the WADA Prohibited List and peptide hormones, growth factors and related substances in S2; both are prohibited at all times, in and out of competition. For several muscle peptides the anti-doping detection literature is larger than the clinical efficacy literature — more scientific work has gone into catching them than into establishing what they do.

Why do peptides get described as a milder alternative?

Because the mechanism sounds milder. A secretagogue asks the pituitary to release more of a hormone the body already makes, which is a gentler-sounding proposition than supplying a synthetic androgen from outside. The physiology behind that framing is real. What does not follow from it is any statement about how much muscle results, because the trials that would establish that were never run — and the mechanism-sounds-mild argument would apply equally to a compound that does nothing.

Is testosterone replacement the same thing as this comparison?

No, and conflating them is the most common error in this debate. The 1996 trial that produced 6.1 kg of fat-free mass used weekly supraphysiologic injections — far above what the body makes or what replacement therapy targets. Prescribed replacement in a man with clinically low testosterone is a different intervention with different expected effects. This page compares what athletes actually take, not what endocrinologists prescribe.

What would settle the question?

A randomised, placebo-controlled trial of a muscle peptide in healthy trained adults on a supervised resistance-training programme, with fat-free mass and one-repetition-maximum strength as endpoints, running for at least twelve weeks. That is the exact design used to test testosterone thirty years ago. It has never been registered for any peptide sold for muscle growth, and nothing on ClinicalTrials.gov currently proposes one.