How to research a peptide stack before you combine anything

A "stack" is two or more compounds taken over the same period. Online, stacks are shared as recipes, with names, doses and schedules. What is almost never shared is the evidence that the combination has been tested, because for nearly every stack people discuss, it has not. This page is not an argument for or against combining anything. It is a method for finding out what is actually known before a clinician and you make that call.

Step 1: evidence for each compound on its own

Start with each compound separately and place it on a ladder. Most research peptides sit on the lower rungs.

For example, BPC-157 has 35 preclinical studies and one clinical study in a 2025 systematic review, plus a two-person safety pilot; its first randomised trial is still recruiting. Semaglutide has phase 3 trials in tens of thousands of people. Treating those as the same kind of "researched" is the first mistake. The peptide library grades each compound on this ladder, and how peptide research works explains the stages in more depth.

Step 2: evidence for the combination

Then ask whether anyone has tested the compounds together. For research peptides the answer is almost always no. The popular BPC-157 and TB-500 combination has, as its entire human dataset, four patients in a 2021 retrospective knee pain review who received BPC-157 with thymosin beta-4. Combinations that drug companies pursue, such as the amylin and GLP-1 pairing of cagrilintide with semaglutide, are developed and tested as combinations in their own trials. That is the standard. A combination assembled from separate compounds has none of that evidence by default.

"Each is safe alone, so together is fine" does not follow. It is the reason drug interaction sections exist.

Step 3: how compounds can interact

• Overlapping effects. Two compounds acting on the same pathway can add up. Several peptides sold for recovery and for body composition act on the growth axis; several sold for metabolism lower glucose or appetite.
• One changes how the other is absorbed. GLP-1 drugs slow stomach emptying, and their labels warn that this can affect absorption of oral medicines taken at the same time.
• Shared side effects. Nausea, low blood pressure or injection site reactions from two products are harder to attribute and can compound.
• Cumulative product risk. The FDA's stated concerns about many research peptides are immune reactions, impurities and poorly characterised ingredients. Each additional unregulated product adds its own share of that uncertainty. See sourcing, purity and testing.
• Attribution. If you start three things in one week and feel better or worse, you cannot know which did it.

Step 4: reading the primary literature

Summaries, including this site's, are a starting point. When a claim matters, find the paper on PubMed and check:

• Who was studied. Rats, healthy young men, older women with a disease? Results rarely transfer across all three.
• How many. A trial of 12 can find large effects and miss everything else.
• Was there a control, and was it blinded? Pain, energy and sleep are especially prone to placebo effects.
• What was measured. A blood level or a surrogate marker is not the same as an outcome you would feel. Elamipretide's approval, for example, rests on a muscle strength surrogate that the FDA requires to be confirmed.
• Was the endpoint pre-specified, and was it met? A trial that missed its primary endpoint and highlights a secondary one is a negative trial.
• Who did it and who funded it. Not a reason to dismiss, but a reason to look for independent replication. A literature mostly from one lab is weaker than it looks.
• Is it registered? ClinicalTrials.gov shows trials that were planned, including ones that never reported.

Step 5: questions to bring to a clinician

• What is the evidence for this compound in someone like me, and for this combination?
• How does it interact with my current medicines and conditions?
• What baseline tests do you want before I change anything?
• What side effects would mean stop, and what would mean call you?
• What would we expect to see, by when, if it is working? And if we see nothing, when do we stop?
• If I take part in sport, is it on the WADA prohibited list?

Step 6: baseline before you begin

Labs drawn before any change are the reference for everything after. A sensible general panel, decided with your clinician, often includes a lipid panel or ApoB, HbA1c and fasting glucose, kidney and liver function, a full blood count and any marker relevant to the compound's mechanism, such as IGF-1 for anything acting on the growth axis. Add the non-lab baselines too: weight, waist, resting heart rate, sleep, and whatever symptom you are trying to change, scored the same way every day for at least two weeks. Our bloodwork guide covers what each marker means.

Step 7: one change at a time

This is the single most useful rule, and the one stacks break by design. Change one thing, hold everything else steady, and give it long enough to show an effect before adding anything. If something goes wrong, you know what caused it. If something goes right, you know what to keep. Write down the start date, the product and its lot number, and every side effect with the day it happened.

For any compound without an established human dose, there is no number we can give you, and a number circulating online is not evidence. That is a conversation for your prescriber.

Where the app fits

Peptide AI was built for exactly this kind of record. The library shows how much research sits behind each compound. The protocol planner lays changes out one at a time, the dose log and vial tracker record what you took and from which lot, and bloodwork tracking puts your labs on the same timeline. Browse every compound on compounds.