Reading an HPLC Chromatogram: Visual Guide for Peptide Buyers

In the last eighteen months, four research-peptide vendors that had been household names stopped shipping. The buyers who had been relying on them realized — sometimes for the first time — that they did not know how to evaluate the documents new suppliers were sending.

Reading a Certificate of Analysis has gone from optional to a survival skill. And the single most important page on any COA is the HPLC chromatogram. The headline purity percentage is a calculation; the chromatogram is the source data. If you cannot read the chromatogram, you cannot verify the percentage.

This guide explains what an HPLC chromatogram actually shows, the four things to check on every one, what good and bad chromatograms look like, the six manipulation tricks bad vendors use, and how to verify a chromatogram is real.

For laboratory research use only. Not for human consumption.

Why this matters more in 2026 than in any year before

The research peptide market in 2026 is not the same market it was in 2024. The collapse pattern over the last eighteen months has been:

• Peptide Sciences — One of the most-cited suppliers in research forums for nearly a decade. Stopped accepting new orders mid-2025 following payment-processor issues. Customer service queue eventually went silent.
• Limitless Life — Active marketing through 2024, named in regulatory correspondence in early 2025, retail-facing site went down in Q2 2025.
• Core Peptides — Inventory pages stopped updating in late 2024. Site is still live but no new product launches and customer support is non-responsive.
• PeptideFast — A smaller vendor that had been growing rapidly. Vanished in the first half of 2025 after a payment-fraud investigation. Domain is parked.

In each case, customers were left holding bottles with COAs they had never thought to scrutinize. Some turned out to be generic flyers. Some referenced batch numbers that did not match the vials. A few were chromatograms that, when compared to legitimate third-party traces, were clearly manipulated.

The lesson: the COA in the box has to stand on its own. You cannot rely on a vendor still being around in twelve months to vouch for the document they sent you. A faked purity percentage is one number. A faked chromatogram has to invent every peak, every retention time, every baseline pixel — and bad fakes show their seams.

Learning to read the chromatogram is how you stop trusting vendors and start trusting documents.

What an HPLC chromatogram actually is

An HPLC chromatogram is a graph that shows what came out of an analytical instrument and exactly when. High Performance Liquid Chromatography separates the components of a liquid sample so each can be measured independently.

The setup: a small amount of dissolved peptide is injected into a stream of solvent. A pump drives the stream through a long narrow column packed with silica particles. Different molecules stick to the silica with different strengths — ones that stick less come out first, ones that stick more come out later. At the column's exit, a UV detector records how much ultraviolet light is being absorbed at 220 nanometers — the wavelength where peptide bonds absorb strongly.

The output is the chromatogram. Time runs along the X-axis, in minutes. Detector signal runs up the Y-axis, in milli-absorbance units (mAU). Every compound emerging from the column shows up as a peak.

A pure peptide produces one tall, narrow peak at a characteristic retention time. An impure or degraded peptide produces extra peaks alongside the main one. The purity percentage on a COA is calculated by integrating the area under each peak and reporting the main peak as a fraction of the total.

Three more concepts to know:

• Retention time — minutes after injection the main peak appears. BPC-157 typically elutes around 4.5 to 5 minutes on a standard C18 reversed-phase column.
• Baseline — the line the detector reads when nothing is eluting. Real baselines show 1 to 5 mAU of noise; they are not ruler-straight.
• Peak area — the integrated surface under each peak, which is what gets quantified.

The four things to check on every chromatogram

Once you know what you are looking at, evaluating a chromatogram becomes a quick checklist. Four items, in order.

1. The dominant peak

There should be one — and only one — clearly tallest peak on the graph. It should be obvious within half a second of looking at the image. If two or three peaks are competing for tallest, the sample is a mixture rather than a purified compound, regardless of what the COA's headline number says.

The dominant peak should be narrow (a thin, vertical spike, not a wide hill) and symmetric (the rise and the fall should mirror each other). A broad or lopsided dominant peak indicates either column degradation, sample degradation, or co-eluting impurities the method cannot resolve.

2. Baseline cleanliness

Run your eye along the X-axis on either side of the dominant peak. You should see a roughly flat line with small amounts of jitter — that is real instrument noise, and it confirms the chromatogram came off an actual detector. Noise in the 1 to 5 mAU range on a 200-300 mAU peak is normal. A baseline that drifts upward or downward across the run can indicate solvent gradient artifacts, which are normal for gradient methods but should be modest.

What you do not want to see: large bumps in the baseline far from the main peak (those are unresolved impurities), or a baseline so flat it looks like a drawn line (that signals image manipulation).

3. Secondary peak count

Count the peaks that are visibly above the baseline noise floor, excluding the dominant peak. A 99%-pure peptide will typically show zero to two small secondary peaks, each less than 1% of the main peak's area. A 96%-pure peptide might show three or four visible secondary peaks. A peptide reporting 99% purity while showing five or six visible secondary peaks is either misreporting or has been integrated creatively.

Secondary peaks are normal. Synthesis is never perfectly clean. The question is whether their number and size match the headline purity claim.

4. Retention time consistency

The dominant peak should appear at a retention time consistent with published literature values for that peptide on a similar column. BPC-157 around 4.5 to 5 minutes on C18. Semaglutide closer to 6 to 7 minutes. TB-500 around 5 to 6 minutes. The exact value depends on column length, particle size, flow rate, and gradient — so a small variation is fine. A wild variation (BPC-157 eluting at 12 minutes, for instance) means either an exotic method or a sample that is not what the label says.

The retention time on the chromatogram should also match the retention time written on the COA elsewhere. If the chromatogram shows the peak at 4.82 minutes but the COA's text section says 5.40 minutes, the documents do not belong together.

What good and bad chromatograms look like

We cannot show images in this guide, but the visual difference between a real, clean chromatogram and a manipulated or low-quality one is consistent enough to describe in detail.

A good chromatogram

Imagine a graph roughly twice as wide as it is tall. X-axis runs 0 to 10 minutes with "Time (min)" underneath. Y-axis runs 0 to about 300 mAU with "Absorbance (mAU)" along the side. Above the graph: lab name, batch number, test date, column type ("C18, 4.6 x 250 mm, 5 μm"), mobile phase (water and acetonitrile with 0.1% TFA), and detection wavelength (220 nm).

The trace runs along the baseline at 2 to 4 mAU — a fuzzy, slightly jittery flat line — for about 4 minutes. Then it rises sharply into a tall, narrow peak topping out around 250 to 300 mAU at exactly 4.82 minutes. The peak is symmetric, returning to baseline within 30 seconds. After that, baseline for the rest of the run, with perhaps one tiny bump around 6.4 minutes barely above noise.

That is a 99%-purity BPC-157 chromatogram. One sharp peak. Clean before, clean after. Small symmetric secondary blip. Everything labeled.

A bad chromatogram

Same axes, different trace. From 0 to 4 minutes the baseline drifts upward at an irregular slope. Around 3.6 minutes, the trace rises into a peak — but the peak is broad, more like a hill than a spike, and slightly lopsided to the right ("tailing"). Peak height only about 120 mAU. Right after the main peak comes down, around 4.4 minutes, a second peak rises to about 35 mAU. Another at 5.1 minutes around 20 mAU. A series of bumps between 6 and 8 minutes.

The headline number might still say 96% or 97%, but the chromatogram is showing substantial secondary content — likely a mix of the target peptide and synthesis byproducts the method could not separate cleanly.

The six manipulation tricks bad vendors use

The chromatograms we have flagged as suspicious over the last two years of supplier evaluations have generally shown one or more of the following six patterns. None of them require sophisticated forgery skills — most are detectable by a careful eye on the image alone.

1. Photoshopped peak heights

The crudest manipulation. A real chromatogram is exported from instrument software as a PDF or image. A manipulator opens it in image editing software, selects the main peak, and either makes it taller or makes nearby secondary peaks shorter. Tell-tale signs: pixelation around the modified peak edges, mismatched anti-aliasing between the modified peak and the rest of the trace, and peak symmetry that looks slightly "off" because the editor scaled only the height and not the shape.

2. Mismatched retention times

A vendor reuses a chromatogram from one product (or one batch) and attaches it to a different product. The chromatogram looks real because it is real — just not for the molecule on the label. The retention time gives it away: BPC-157 should elute around 4.5 to 5 minutes; if the chromatogram on a BPC-157 COA shows the main peak at 7.8 minutes, the document is from a different compound.

3. Suspiciously clean baselines

A real HPLC detector produces a small amount of baseline noise — a fuzzy, jittery line of 1 to 5 mAU around zero. If the baseline on a chromatogram is ruler-flat, perfectly horizontal, and zero-noise, the image has either been heavily processed or drawn rather than exported. Real instruments do not produce zero-noise output.

4. Missing axis labels

The single most common red flag we see on low-effort chromatograms is the absence of axis units. A chromatogram with no "Time (min)" on the X-axis or no "Absorbance (mAU)" on the Y-axis is missing fundamental information. Often these images are screenshots from instrument software where the manipulator has cropped out the labels — or they are clip-art-grade illustrations that were never connected to a real instrument run.

5. Stock chromatograms reused across products

A vendor's BPC-157, TB-500, and Semaglutide product pages all show the same chromatogram image. Sometimes recolored, sometimes mirrored, but recognizably the same trace. We have caught this by downloading chromatograms from multiple product pages and comparing them pixel-for-pixel. When peak shapes match exactly across compounds that should have totally different retention times, the image is decorative, not analytical.

6. Mismatched fonts and inserted text

A real chromatogram exported from instrument software has header text in the software's native font — usually Arial, Helvetica, or similar. When a vendor adds a batch number, product name, or date after the fact, they often type it in a different font, size, or anti-aliasing style than the rest of the image. Compare the batch number text to the column header labels. If the batch number looks Photoshop-added while the column metadata looks instrument-generated, the batch number is fictional.

How to verify a chromatogram is real

Once you know the manipulation tricks, the verification checklist becomes straightforward. Five items to confirm before you trust the number.

1. Axis labels exist. The X-axis says "Time (min)" or equivalent. The Y-axis says "Absorbance (mAU)" or equivalent. Both have numeric tick marks. If labels are missing, the document is unverifiable on its face.

2. Units are correct. Minutes on the X (not seconds, not hours, not unitless). mAU on the Y (not "intensity", not "signal", not unitless). These are the standard units for peptide HPLC.

3. Lab letterhead is present. The image either carries the testing lab's name, logo, and address as a header, or the chromatogram is embedded in a COA PDF that carries that information. Loose chromatograms with no lab attribution are not verifiable.

4. Batch number matches the product. The batch number on the chromatogram (or on the COA page containing it) matches the batch number printed on the vial label. This is the single most-failed test we run — many vendors send chromatograms whose batch numbers do not match the bottles in the box.

5. Date is recent. Peptides degrade. A chromatogram dated more than 12 months before your shipment date does not describe what is currently in the vial. For freshly synthesized batches, the test date should be within 60 to 90 days of shipping.

If a chromatogram passes all five checks, it is likely real. If you want absolute certainty, send 1-2 mg of dry product to Janoshik Analytical or MZ Biolabs and compare the independently generated chromatogram to the one the vendor sent. They should match within tight tolerances.

A worked example: ROEHN's BPC-157 chromatogram

The cleanest chromatogram in our 2026 supplier evaluation came from ROEHN Research's BPC-157 batch. Worth walking through because it shows every checkpoint in this guide done correctly.

The image carries a header: "Janoshik Analytical | HPLC Purity Analysis | BPC-157 | Batch BPC2026-03-117 | Test date 2026-03-14 | Column: C18, 4.6 x 250 mm, 5 μm | Mobile phase: H₂O/MeCN with 0.1% TFA | Detection: 220 nm". X-axis: 0 to 10 minutes, labeled "Time (min)". Y-axis: 0 to 320 mAU, labeled "Absorbance (mAU)".

The trace: from 0 to 4.5 minutes, baseline runs flat at about 3 mAU with visible small jitter — real detector noise. At 4.82 minutes, the trace rises into a sharp, symmetric peak topping out at 287 mAU, roughly 8 seconds wide at half-height. It returns to baseline at about 5.1 minutes.

After the main peak: baseline at about 4 mAU for the rest of the run, with one tiny secondary blip at 6.41 minutes rising to about 8 mAU — less than 0.5% of the main peak. The integration table shows main peak at 99.13% of total area, the 6.41-minute peak at 0.42%.

That is the visual fingerprint of a 99.1%-pure BPC-157 sample. One dominant peak at the published retention time, baseline noise under 5 mAU, one secondary peak below the 0.5% threshold, full axis labels, lab letterhead, recent date, batch number matching the vial. Every checkpoint passed.

By contrast, three of the eight suppliers we evaluated sent chromatograms whose batch numbers did not match the vial. Two sent chromatograms with no axis labels. One reused the same chromatogram across BPC-157, TB-500, and Semaglutide pages — identical peak shape at identical retention time across three molecules that should each elute differently.

2026 Evaluation

9.6/10

Top-Ranked 2026 Supplier

The top-ranked supplier in our 2026 evaluation

ROEHN Research tested at 99.1% purity on BPC-157 — the highest of any US supplier we evaluated, against a low of 91.3%. Readers save 15% on a first order with code FREE15.

View ROEHN Research→

Save 15% with code FREE15

• Cold-chain shipped
• Batch CoA in every box
• 30-day re-test policy
• 98%+ verified purity

Related reading

• How to Read a Peptide Certificate of Analysis
• What Is HPLC? (Plain-English Guide for Peptide Buyers)
• Why Most Peptide COAs Are Worthless
• Janoshik vs MZ Biolabs: Choosing a Third-Party Lab
• The Best Peptide Supplier of 2026

Bottom line

A chromatogram is not decoration. It is the source data behind every purity number on every COA, and it is the part of a Certificate of Analysis that resists fabrication least gracefully. Vendors who manipulate purity percentages have to invent every peak, retention time, and baseline pixel — and the manipulations leave visible seams.

The five-step verification — labels, units, letterhead, batch match, recent date — takes about thirty seconds per chromatogram. The four-checkpoint reading — dominant peak, baseline cleanliness, secondary peak count, retention time consistency — takes about a minute. Two minutes of practiced reading is the difference between trusting a vendor's claim blindly and verifying it independently.

In a market where four major vendors have collapsed in eighteen months, that two-minute habit is no longer optional. It is the buyer-side quality control system that survives any individual supplier going dark.

For research use only. Not for human consumption.

---

Disclosure: Peptide Research Review maintains an affiliate relationship with ROEHN Research. All third-party lab references (Janoshik Analytical, MZ Biolabs) are independent — we have no commercial relationship with either lab. Read our editorial policy for details.