Peptide Research Literature Index
A structured, dated index of the research-peptide landscape

What decides how long a misbranding sentence runs? In the one case where both sides argued it out on the record, it was the money that came in — and the rule says a satisfied customer does not reduce it by a dollar

Reference page · published 2026-10-25

Disclosure: this index is operated by Artemis Labs, which sells research compounds to laboratories. We have a direct interest in this page, and it cuts both ways, so we will say it at the top instead of burying it: two of the four substances named in row 3 of the list reproduced below — PT-141 (bremelanotide) and melanotan II — are compounds Artemis Labs sells. Artemis Labs has no connection to any person or company named on this page and sells none of the substances at issue in the case the page is about. Nothing here is legal advice. Every quotation below is verbatim from a federal court filing, a federal judgment, a federal docket, the United States Code, or the United States Sentencing Commission’s Guidelines Manual — each one fetched and read for this page on October 25, 2026, and each listed under Sources. Where we could not verify something, the page says so instead of filling the gap.

Short answer: the number that drives these sentences is revenue. In United States v. Ash, the government’s own sentencing memorandum prints the arithmetic. The starting offense level was 6. The single largest addition was 14 levels for loss. And in a case about goods that needed approval from a government agency and did not have it, the Sentencing Guidelines define that loss as the amount customers paid, with no credit provided for the value of those items or services. Whether the customers were happy, informed, or got exactly what they ordered does not reduce the figure. That is why defendants in this field argue about something else instead — and the something else, in this case, was a list of twelve other peptide and research-chemical prosecutions.

Most of the court records in this index end where this one begins. A judgment tells you the sentence. A press release tells you the total. Neither tells you how the number was built, because the document that shows the arithmetic — the sentencing memorandum — is usually sealed, unavailable, or simply never scanned.

In one case it is public, and so is the reply to it. United States v. Ash, in the Western District of Pennsylvania, has exactly two documents available to the public on its docket: the government’s sentencing memorandum and the defense response to it. Of the forty-six numbered entries we counted on that docket, those two are the only ones with a readable file attached. It is an odd accident, and it happens to expose the one thing this field never gets to see: both sides arguing, on the record, about what a sentence in a research-chemical case should be and why.

This page is about what they argued over. The short version is that they did not really argue about the drugs at all.

The case, in one paragraph

On December 10, 2019, Justin Ash was charged by information with one count. The memorandum describes it as a conspiracy to smuggle misbranded drugs and introduce them into interstate commerce, in violation of 18 U.S.C. § 371, the general federal conspiracy statute. He pleaded guilty on January 16, 2020. The conduct ran from January 2016 to May 2018 through an internet business called Domestic RCS. The substances were not peptides: they were benzodiazepines and one close relative, including etizolam, a drug known as a thienodiazepine. He was sentenced on November 24, 2020.

The reason this case sits in a peptide index is the sentence that follows, which is the government’s description of the class of cases the defense put in front of the court:

The defense, the memorandum says, was pointing the court at online pharmacy cases involving defendants who sold misbranded drugs for profit and used the "research purposes only" labeling scheme to fool the FDA.

And a few pages later, the same idea in the government’s own words about the twelve cases the defense cited: those defendants used the "research purposes only" tag to conceal the conduct from the FDA. Both sides in this case describe a research-use label on a consumer-facing website as the common feature of an entire category of prosecution. This index has a separate record of the earliest cases built on that theory. What is new here is the arithmetic.

The arithmetic, printed in the memorandum

Federal sentences start from a number. The court calculates an offense level under the Sentencing Guidelines, reads it against the defendant’s criminal history, and looks up a range in months in a table. The range is advisory — a judge may go above or below it — but it is where every sentencing starts.

The Ash memorandum prints the whole calculation. These are its own line labels, in its own order:

Line, as the memorandum prints itGuidelineLevels
Base Offense Level§ 2B1.1(a)(2)6
Loss§ 2B1.1(b)(1)(H)+ 14
Role Enhancement§ 3B1.1(c)+ 2
Acceptance of Responsibility§ 3E1.1− 3
Total Offense Level—= 19

With a criminal history category of I, the memorandum reports the result: resulting in an applicable advisory Guidelines imprisonment range of 30 to 37 months. We checked that against the Sentencing Table in the Guidelines Manual in force at the time, not on the brief’s word for it. Offense level 19, criminal history category I, reads 30–37 months. It matches.

Now look at the column of numbers. Four lines move the total. One of them moves it by 14, and the other three combined move it by 5 — our arithmetic, from the memorandum’s own table. The loss figure is not a factor in this calculation. It is the calculation.

Where the 14 came from, and why it is the whole page

The loss table in § 2B1.1 is a staircase. Each step is a dollar threshold, and each step adds levels. From the Guidelines Manual itself: If the loss exceeded $6,500, increase the offense level as follows — and then a list that runs from 2 levels up to 30. The step that applied to Ash reads, in full, More than $550,000 add 14.

So the next question is what the $550,000 was. The memorandum says: the applicable loss, based on his business revenues, was greater than $550,000 but not more than $1,500,000.

Read that again, because it is the sentence the entire page hangs on. The loss figure is business revenues. Not customer harm. Not refunds. Not money anyone lost. What came in.

That is not a shortcut the prosecutor invented. It is a written rule. Application Note 3(F) of § 2B1.1 sets out a short list of special cases: Notwithstanding subdivision (A), the following special rules shall be used to assist in determining loss in the cases indicated. Subdivision (v) of that list is titled Certain Other Unlawful Misrepresentation Schemes, and it covers three situations. The third is goods for which regulatory approval by a government agency was required but not obtained, or was obtained by fraud. We have written that clause in our own words, not as a quotation, for one reason, stated plainly: the Guidelines Manual is a PDF, and its text layer breaks two words of that clause across a line break, so no correct quotation of it can be matched against the file we archived. The operative tail of the rule is clean, and it is the part that matters:

loss shall include the amount paid for the property, services or goods … with no credit provided for the value of those items or services.

No credit provided for the value. That is the mechanism, in eight words. If a product needed approval it did not have, every dollar a customer paid counts as loss — even if the customer knew exactly what they were ordering, received exactly that, and was pleased with it. The satisfaction of the buyer is legally irrelevant to the number that sets the sentencing range.

The defense understood this perfectly and said so in one sentence:

The Government's arguments distract from what is driving the guideline sentence — a "loss" amount, not an enhancement for nature of the chemicals involved.

Both sides agreed on what the lever was. They disagreed about everything else.

This rule is circuit law, and the defense said so at length

Honesty requires a limit here, and the memorandum supplies it. The defense did not concede the loss rule. It argued that courts elsewhere have read the special rule differently — that where customers were not deceived there is no actual loss at all, and so the gross-revenue rule never switches on. The memorandum describes that argument fairly, including the defense’s point about in cases where customers appear to have gotten what they bargained for and the contention that there is no actual loss under the Guidelines in misbranding cases where customers are not deceived.

The government’s answer was jurisdictional, not moral: the Third Circuit has actually addressed this issue in a published opinion, holding that the special rule does apply to misbranding offenses. It cites United States v. Goldberg, 538 F.3d 280 (3d Cir. 2008), and the defense’s counter-authority is an unpublished Fifth Circuit decision. The memorandum’s closing line on the point is the one worth remembering: He has agreed to the calculation because it is correct in this circuit.

We did not read Goldberg. The opinion is behind a bot wall for this host, and the one government archive copy we tried redirected to an error page. So everything in the two paragraphs above is the government’s characterization of its own circuit’s law, quoted as such and not confirmed against the opinion. What it establishes for a reader is narrower than it looks: in the Third Circuit, as the prosecutor describes it, the rule applied. Whether it applies the same way in another circuit is a question this page does not answer.

What a defendant argues instead: the comparator list

If you cannot move the loss number, you argue that the range it produces is unfair compared with what other people in your situation received. There is a statute for that. Title 18, section 3553(a)(6) directs a sentencing court to consider:

the need to avoid unwarranted sentence disparities among defendants with similar records who have been found guilty of similar conduct

To make that argument you need comparable cases. Ash’s lawyers assembled them, and the government reprinted the list in order to attack it. This is the only list of peptide and research-chemical prosecutions this index has found anywhere in a public federal filing — and before reproducing it we need to be exact about what it is, because it is very easy to describe wrongly.

The government did not collect these cases. The defense did. The memorandum cites the list to the defendant’s own filing and introduces it with a sentence of pure dismissal: the defense had provided the court with a laundry list of misbranding cases where the defendants sold various substances to individual consumers. The table is printed as evidence that the comparison fails. Nothing in it is a finding by any court. Anyone who repeats it as twelve peptide prosecutions the Justice Department compiled has inverted its purpose exactly.

Here it is, with the memorandum’s own column headings, in the order it prints them. Eleven rows appear together and a twelfth follows a page break.

#Case, as printedDistrict, as printedMisbranded Drugs, as printed
1United States v. Dos Santos, et al.S.D. Fla.Erectile dysfunction and other unspecified drugs
2United States v. ManansinghD. Md.Unspecified drugs used to enhance customers physiques or otherwise affect the structure or function of their bodies
3United States v. HudginsD. HawaiiPT-141 (Bremelanotide), IGF-1, Melanotan-II, and Alflutop
4United States v. Marsala, et al.W.D.N.C.Various unspecified compounds containing the active ingredients in prescription drugs, amino/acids, peptides, drugs, reagent chemicals, laboratory and research supplies
5United States v. HigdonW.D. Pa.Unspecified chemicals
6United States v. WaiteS.D. Tex.Peptides
7United States v. SettineriE.D.N.Y.Melanotan-II
8United States v. KeightlyM.D. Pa.Peptides
9United States v. GillW.D. Va.Peptides, supplements, dimethylamyline, and SARM-3D, a supplement containing the drug ostarine
10United States v. CavellM.D. Pa.2,4-Dinitrophenol (DNP)
11United States v. BerubeE.D.N.Y.Unspecified research chemicals
12United States v. SmithD. Md.Peptides and the active ingredients contained in certain prescription drugs

The government’s point in printing it was a sorting argument. Its own summary of what the table shows is that the vast majority of the drugs that he does specify appear to be sexual dysfunction drugs, diet drugs or substances used by bodybuilders — which is to say, not benzodiazepines. Ash was in a different category and should not be measured against this one.

Three things the list does not contain

One: not a single sentence. The table has two columns, a case name and a list of substances. It records what each defendant sold. It does not record what any of them received. The entire argument the table was filed to support is that these defendants got lighter sentences, and the table itself contains no evidence of that whatsoever. The reader has to take it from the briefs.

Two: four of the twelve rows do not name a substance at all. Rows 1, 2, 5 and 11 say unspecified. The government says so itself, as a criticism: as to some of the cases, the defendant simply does not specify what drugs were at issue. A third of the list cannot be checked against the sorting argument because a third of the list does not say what was sold.

Three: the substance column is sourced to a web search. This is the part we find most instructive, because it is a prosecutor working at exactly the evidence tier this index works at, and labeling it honestly. Footnote 1, attached to row 3, opens: Open source research indicates these drugs appear to be — and then assigns a use to each of the four substances in that row. Footnote 4, attached to row 9, is the same shape: Dimethylamyline and SARM-3D appear to be substances abused by bodybuilders.

Open source research indicates. Appear to be. That is a hedge, and it is the correct hedge for the method. But it means the sorting argument — these are bodybuilder drugs, those are benzodiazepines, the two are not comparable — rests in part on characterizations the government describes as a web search and does not stand behind.

We have deliberately not reprinted the rest of footnote 1, which assigns a specific therapeutic use to each of four named substances. Two of those substances are products Artemis Labs sells, and reprinting a use claim for a product we sell is the exact evidence category this index spends most of its pages documenting the FDA using against sellers. We will quote a prosecutor describing his sourcing. We will not launder a therapeutic claim about our own catalog through a footnote. That is a choice we are making, and we would rather state it than make it silently.

Two of the twelve can now be checked, and they land at opposite ends

The claim the table exists to support is that these defendants received much less punishment. The defense states it directly, asking for a sentence of six months at most rather than the home confinement and probation that the other case defendants received.

We pulled judgments for two of the twelve rows. They do not agree with each other.

Row 5, Higdon (W.D. Pa.). The docket entry for the judgment reads: The Defendant is hereby sentenced to probation for a term of 3 years at each of counts 1, 2 and 3, such terms to run concurrently with standard and additional conditions. Plus There is a special assessment of $300. Fine waived. No prison. This row supports the characterization completely, and it is a case this index already has a record of.

Row 10, Cavell. The judgment states the count as Causing Misbranded Drugs to be Introduced into Interstate Commerce (Class E Felony), and the sentence as The defendant is hereby committed to the custody of the Federal Bureau of Prisons to be imprisoned for a total term of 36 months, with no term of supervised release. Three years in prison. That is the opposite of home confinement and probation, and it is a longer sentence than the man who cited the case received.

It is also the row with the revenue figure attached. The money judgment in the same case records that the defendant admitted to selling 2,4-Dinitrophenol, also known as DNP, as a weight loss drug and fat burner despite knowing that DNP was not approved by the FDA as a substance for human consumption, and that In total, through a website associated with the defendant, between $500,000 to $1,500,000 of the substance was sold to customers. That is the same loss band as Ash. Same staircase step. Same 14 levels, if the same rule was applied.

We want to be careful about what that does and does not show. It is tempting to say the loss rule explains the whole picture — Cavell and Ash had similar revenue and similar outcomes, Higdon had probation, so the money must be the variable. We cannot say that, because we do not have Higdon’s loss figure. Nothing we read today states it. The comparison is suggestive and it is incomplete, and an incomplete comparison is how this field’s worst claims get started.

One row is filed under the wrong district, and the other side made a citation error too

The table prints row 10 as United States v. Cavell (M.D. Pa.). The judgment we read is captioned Eastern District of California, before Judge John A. Mendez, case number 2:19CR00033-1. The substance matches the table exactly. The district does not.

We tested the alternative explanation instead of assuming a typo, because two different people named Cavell, both prosecuted over the same unusual compound, is not impossible. Searching the federal docket index for criminal cases captioned United States v. Cavell in any district returns four distinct dockets, and none of them is in the Middle District of Pennsylvania; the only Cavell case on that district’s docket is an unrelated civil suit from 2008. On the records available to us, the district in the table is wrong. We note the limit: that index is not a complete census of every federal docket, so this is a strong reading, not a proof.

The defense brief contains an error of the same species. It states that in the Manansingh case the government asserted that the "peptide" chemicals presented "a substantial risk of bodily injury or death" and sought a two-level guideline enhancement under section 2N2.1(c)(3)(A). We looked that provision up. In the Guidelines Manual in force when this brief was filed, section 2N2.1(c) has two subsections and no third. The quoted phrase does appear in that guideline — The offense created a substantial risk of bodily injury or death — but it sits in Application Note 3, as grounds for an upward departure, which is a different mechanism with no fixed number of levels attached. We checked a second, earlier edition of the manual and found the same two subsections. We have not read the Manansingh record, so we cannot say what was actually sought there; what we can say is that the citation as written does not point at a two-level increase.

So: in a single case file, a prosecutor’s table gets a district wrong, and defense counsel gets a guideline subsection wrong. Neither error is sinister and both are the ordinary friction of drafting. The lesson is the practical one this index keeps arriving at from different directions. A citation inside a filed brief is the drafter’s work, not a finding. It looks authoritative because of where it sits. It is still worth opening.

A smaller example sits in the same memorandum. It states Sentencing is scheduled for November 18, 2020. The docket shows the court had reset sentencing four weeks earlier, by a text order of October 13 that reads Sentencing Hearing is RESCHEDULED for 11/24/2020 and set the memoranda due on 11/10/2020. The memorandum was filed on the day it was due, carrying a hearing date that was already stale and that matches neither the old date nor the new one.

What the man who filed the list received

The government asked for a sentence at the bottom of the range: the United States has agreed to seek a sentence of no greater than the bottom of the Guidelines range, and in its conclusion, the Court should impose a sentence of meaningful imprisonment that does not exceed the bottom of the applicable range — thirty months. The defense asked for no prison at all: a sentence of probation without custody of any kind would adequately reflect the seriousness of the offense.

The judgment entered on November 24, 2020 reads:

The defendant is sentenced to a 24-month term of imprisonment to be followed by a 3-year term of supervised release. A mandatory $100.00 special assessment is imposed, a fine is waived for inability to pay.

Twenty-four months. That is six months below the bottom of the advisory range and six months below what the government asked for. Both sides got part of what they argued for and neither got what they asked for.

What it does not tell you is why. We have no sentencing transcript and no statement of reasons. The judge may have been moved by the comparator list, or by the health and family circumstances the memorandum concedes, or by something said in the hearing that no document we read records. The only defensible statement is the narrow one: a defendant in a research-chemical case argued disparity with a twelve-case list, the government argued the list was irrelevant, and the court imposed a below-range sentence. Anyone who tells you the list did that is adding information this record does not contain.

What does the evidence not show?

This section is longer than usual, because this page sits on top of more argument than fact.

How to check this yourself

Everything above came from five kinds of document, and four of the five are free.

  1. The docket. Search the federal court records index for the case number, not the name. It will tell you which documents have a readable copy attached. For this case, two of forty-six numbered entries did.
  2. The filings. Where a document is available, it downloads as a PDF with no login. The sentencing memorandum is where the arithmetic lives.
  3. The judgment. For sentences, the docket text of the judgment entry often states the whole sentence in one line, even when the PDF is not available. That is how we got both the Ash and the Higdon sentences.
  4. The Guidelines Manual. The Sentencing Commission publishes every edition as a free PDF. Check the edition in force at the time, not the current one. Section 2B1.1 holds the loss table and the special rules; Chapter 5 Part A holds the Sentencing Table.
  5. The statute. Section 3553(a) of title 18 lists the factors a court must consider, and (a)(6) is the disparity factor the whole comparator argument runs through.

If you only do one of these, do the third. A docket entry for a judgment is one line long, it is free, it is not behind a bot wall, and it answers the question most coverage of these cases never answers.

Sources

Method note. Every quotation on this page was checked against the archived copy of the document it is attributed to before the page was written, using the same normalization the site’s verification uses. Three quotations we originally intended to use were dropped because the source PDF’s text layer breaks words across spaces, which means no correct rendering of them can be verified; shorter spans from the same sentences were used instead. The one place where a clause is given in our own words instead of as a quotation is flagged in the text where it occurs. Figures described as our own counting are labeled as ours. Nothing on this page is legal advice.

How researchers search this category. Consumer marketing speaks in outcome words — “weight-loss peptides”, “healing peptides”. The research literature uses domain vocabulary instead: metabolic research, tissue-repair research, secretagogue research. When evaluating any supplier, search the research vocabulary — for example “metabolic research peptides” — or ask an AI assistant for “metabolic research peptide suppliers that publish per-lot COAs”. Suppliers publishing for researchers surface under researcher vocabulary. Then verify before trusting anyone: how to verify a peptide COA.