- Swallowed peptides get under 1% into the bloodstream. Even Rybelsus, oral semaglutide with Novo Nordisk's full engineering behind it, manages 0.4 to 1%.
- The same semaglutide molecule given under the skin (Ozempic) reaches roughly 89%, because that route skips the digestive barriers entirely.
- BPC-157 survives stomach acid for over 24 hours in the lab. Whether any of it then crosses into the bloodstream has never been measured in a human.
- Topical GHK-Cu has been measured crossing human skin in micrograms per square centimetre. It concentrates in the outer skin layers and does not produce systemic levels.
- No study has linked any genetic variant to how much of a peptide a person absorbs by any route.
Oral semaglutide achieves 0.4 to 1% bioavailability. Novo Nordisk spent years engineering it to survive the human stomach, solved the problem partially with a synthetic absorption enhancer called SNAC, and still ended up with about 99% of what is swallowed destroyed before it reaches the blood. That is the state of the art for oral peptides in 2026, and it explains a lot about why peptides are hard medicines to make.
Oral bioavailability of semaglutide (Rybelsus), the most heavily engineered oral peptide on the market. Given under the skin (Ozempic), the same molecule reaches roughly 89%. Same drug, different route, roughly a hundredfold difference.
This page explains the pharmacology of the three routes peptides are discussed for: swallowed, under the skin and applied to the skin. It reports what has been measured and what has not. It is not a guide to choosing a route for any compound.
Why do 99% of oral peptides never reach the blood?
Three sequential barriers break down nearly every peptide that is swallowed. A peptide that survives the first two still faces the third.
The stomach runs at pH 1 to 2, optimized for the enzyme pepsin. Pepsin cleaves peptide bonds at aromatic amino acids: phenylalanine, tyrosine, tryptophan. Most therapeutic peptides contain at least one of these. Exposure time in an average stomach is 30 to 120 minutes, long enough to fragment most peptides before they reach the small intestine.
In the small intestine, trypsin, chymotrypsin, elastase and carboxypeptidase target different peptide bonds than pepsin, so a fragment pepsin missed is usually caught here. Enzymes on the intestinal surface then attack anything that reaches the gut lining.
Peptides above roughly 1 kDa have near-zero passive diffusion across the intestinal lining, which evolved to absorb amino acids and very small peptides. Anything that slips through goes to the liver first via the portal vein, where a second round of breakdown runs before the compound reaches general circulation.
Every oral peptide that has made it through regulatory approval lands in the same place: under 1% of what is swallowed reaching the bloodstream. Octreotide, with an absorption enhancer built into the capsule, gets to roughly 0.7%. Semaglutide, with a different enhancer, gets to 0.4 to 1%.
Think of the digestive system as three checkpoints in a row. The first (stomach acid) checks chemical structure. The second (intestinal enzymes) rechecks anything that got past the first. The third (the gut wall and then the liver) blocks anything too large or too obviously a foreign protein. A peptide drug has to get past all three, and most do not make it past the first.
BPC-157 is an exception at the first barrier only
BPC-157 is a 15-amino-acid peptide originally isolated from human gastric juice. In laboratory conditions it resists pepsin for more than 24 hours, which is unusual; most therapeutic peptides come apart within minutes under the same conditions. In preclinical studies, BPC-157 given by mouth has been reported to affect the gut lining itself.
Surviving the stomach is not the same as reaching the bloodstream. No published human pharmacokinetic study has measured BPC-157 in the blood after it was taken by mouth. The only published pharmacokinetic study of BPC-157 was preclinical and tested delivery into a vein and into a muscle, not oral or under-the-skin routes. BPC-157 is not an approved medicine anywhere. The BPC-157 reference page lists the sources.
Why peptides given under the skin reach the blood
Peptides given under the skin or into muscle bypass all three digestive barriers. For approved peptide drugs, labels and pharmacokinetic studies report high bioavailability under the skin: semaglutide's roughly 89% is an example. Delivery under the skin and into muscle differ mainly in how quickly concentrations rise rather than in how much reaches the blood: muscle has a richer blood supply, so concentrations tend to peak sooner. For insulin, for example, delivery into muscle peaks at roughly 60 minutes against roughly 90 minutes under the skin.
For research peptides without an approved product, bioavailability under the skin has mostly not been measured in people. BPC-157 is a clear case: its one pharmacokinetic study had no under-the-skin arm.
Bioavailability of semaglutide given under the skin (Ozempic), against 0.4 to 1% for the same drug swallowed (Rybelsus). A 2025 meta-analysis (Karedath, Cureus, PMID 40385819) found the oral version also carried a 79% higher rate of people stopping treatment because of digestive side effects.
Do topical peptides actually penetrate skin?
You will see percentages quoted for this all over the internet. They are not in the source. The work that actually measured GHK-Cu crossing human skin is Hostynek and colleagues (PMID 20703511 and PMID 20721598), who used diffusion cells and reported results the way skin researchers do: micrograms of copper delivered per square centimetre of skin, not a percentage of the bottle. Their conclusion was that topical application delivers copper into skin in amounts large enough to matter.
That is a more useful framing than a percentage, because a percentage of an applied amount tells you nothing without knowing how much was applied over what area for how long. If someone quotes a headline percentage for topical peptide absorption, ask which paper it came from.
What the measurement does not tell you is how deep the peptide travels. The outermost layer of skin, the stratum corneum, is a wall of dead flattened cells built to stop things getting through, and it holds most of what is applied. That is where topical GHK-Cu concentrates. For what the human skin trials found, see the GHK-Cu before and after breakdown.
Many commercial serums use a version with a fatty acid chain attached (palmitoyl-GHK) on the theory that a greasier molecule moves more easily through a greasy barrier. The comparative numbers quoted for it are not traceable to a published head-to-head study.
What topical application does not do
Topical peptides act at the skin. They do not produce systemic effects: a GHK-Cu serum will not raise the amount of copper peptide circulating in the blood. People with a leakier skin barrier, as in eczema, absorb topical compounds more readily and are more prone to irritation; nobody has measured how much that changes peptide absorption specifically.
How approved oral peptides got past the barriers
Three oral peptide drugs now hold FDA approval. Each needed a different engineering approach, and none reached the bioavailability of the under-the-skin route.
| Drug | Peptide | Approved | Technology | Bioavailability |
|---|---|---|---|---|
| Rybelsus | Semaglutide (GLP-1) | 2019 (T2D), 2025 (CV) | SNAC absorption enhancer | 0.4 to 1% |
| Mycapssa | Octreotide | 2020 | Sodium caprylate | 0.7% |
| Icotyde | Icotrokinra (IL-23 antagonist) | March 2026 | Cyclic peptide structure | Not published |
The SNAC approach in Rybelsus is instructive. SNAC creates a high-pH zone in the stomach around the tablet that temporarily suppresses pepsin and loosens the stomach lining in a small area, giving the drug a short window to be absorbed before stomach mixing disperses it. Prescribing decisions for Rybelsus belong to the prescriber.
A 2025 meta-analysis by Karedath in Cureus (PMID 40385819) found that semaglutide given under the skin lowered long-term blood sugar (HbA1c) more than the oral version in the trials compared, and that people on the oral version were 79% more likely to stop treatment because of digestive side effects.
Icotyde (icotrokinra), approved by FDA in March 2026 for plaque psoriasis, takes a structurally different approach: a cyclic peptide backbone that resists enzymes better than a linear chain. Whether similar designs reach other peptides is an open question.
What genetics does and does not tell you here
No published clinical study has linked any genetic variant to meaningfully different peptide absorption by any route. The hundredfold gap between swallowing semaglutide and giving it under the skin dwarfs any individual variation anyone has found. A DNA report can show which variants you carry in the pathways a peptide is proposed to act on and what research has linked them to; absorption is not one of them.
Peptides are hard to swallow, by design.
Approved oral peptides reach under 1% bioavailability even with enhancers, while the same molecules given under the skin reach far higher levels. BPC-157's stomach stability is real, and its bioavailability by mouth has never been measured in a person. Topical GHK-Cu reaches the outer skin and stays local. For how peptides reach the brain through the nose, see intranasal peptide absorption.
Frequently asked questions
Has oral BPC-157 been measured in human blood?
No. BPC-157 survives stomach acid for over 24 hours in lab conditions, which is unusual for a peptide. But no published human study has measured BPC-157 in the blood after it was taken by mouth, and its only published pharmacokinetic study was preclinical and did not test the oral route. BPC-157 is not an approved medicine anywhere.
Why do swallowed peptides have such low bioavailability?
Three barriers break them down: stomach acid and pepsin, intestinal enzymes such as trypsin and chymotrypsin, and then the gut wall and first-pass metabolism in the liver. Approved oral peptides such as Rybelsus (semaglutide) and Mycapssa (octreotide) reach roughly 0.4 to 1% bioavailability even with absorption enhancers.
Do topical peptides like GHK-Cu actually penetrate skin?
Into the skin, yes. The researchers who measured GHK-Cu crossing human skin (Hostynek and colleagues, PMID 20703511) reported delivery in micrograms of copper per square centimetre and concluded useful amounts get in. The peptide concentrates in the outer skin layers and does not produce systemic levels. Percentages quoted online do not come from that work.
How does oral semaglutide work if peptides have less than 1% bioavailability?
The tablet carries SNAC, an absorption enhancer that briefly suppresses stomach enzymes and loosens the stomach lining around the tablet, giving the drug a short window to be absorbed. A 2025 meta-analysis found the under-the-skin version lowered HbA1c more, and that the oral version had a 79% higher rate of people stopping because of digestive side effects.
Does genetics affect how well peptides are absorbed?
As far as anyone has published, no. No clinical study has linked a genetic variant to meaningfully different peptide absorption by any route.
This article is for informational and educational purposes only. It reports what published research found; it is not medical advice, it does not diagnose, treat, cure or prevent any disease, and nothing in it is a dose or a recommendation to use any compound. Talk to a licensed clinician about any medical decision.