Doping control in commercial combat sports: the money, the laboratory, and a contract annex

In October 2026, the United States Anti-Doping Agency (USADA) announced it was ending its partnership with the UFC after more than eight years, effective 1 January 2026. Three months later, the anti-doping programme of the largest mixed martial arts promotion moved to Drug Free Sport International. Most coverage over the following fortnight revolved around a handful of fighter names and the question of who would be allowed back into the cage. One line in the USADA statement was almost never quoted: the condition for competing is six continuous months inside the testing pool, and any private arrangement around that window changes the value of the entire data set behind it.

I logged that date into a dedicated column of my personal spreadsheet, where I keep doping-control records from combat sports going back to 2026. The adjacent columns hold the effective date of each rule set, the sample collection agency, the laboratory, and the arbitration venue. Those four fields, combined, say more than any allegation.

The arena is clean. The dressing room is not.

A decade of transferred authority

Before July 2026, major mixed martial arts promotions in the United States tested under state athletic commission rules. Every state had its own thresholds, its own collection process, its own calendar. When the UFC signed with USADA, the model changed axis: year-round out-of-competition testing, whereabouts obligations, samples shipped to laboratories accredited by the World Anti-Doping Agency (WADA), and sanctions decided by an independent arbitrator rather than the promoter.

The measurable effect followed. In 2026, Anderson Silva was found to have used drostanolone. In 2026, Brock Lesnar tested positive for clomiphene and hydroxy-clomiphene weeks after returning at UFC 200. In 2026, Jon Jones was sanctioned for turinabol after an out-of-competition test ahead of UFC 214. Those three names are not evidence that doping increased. They are evidence that a system could see what it previously could not.

Conor McGregor was the name most discussed in that period, but the technical question sits elsewhere: six continuous months in the testing pool is what makes one result comparable to the same athlete's results at another point in time. Without that anchor, a clean sample is just a clean sample.

The part rarely written about is the contract. USADA is not a state regulator. It is a non-profit paid by the UFC for a service. WADA does not run this promotion. The UFC sits outside the World Anti-Doping Code, as do the NBA, the NFL and the NHL. That segment of professional sport runs on a private model: the promoter writes the rules, hires the enforcer, pays the laboratory and appoints the arbitrator.

Olympic boxing travelled a different route and arrived at the same question. The International Boxing Association, formerly AIBA and later IBA, was suspended by the International Olympic Committee (IOC) in 2026 following judging controversies at Rio 2026 and governance and finance concerns. On 22 June 2026, the IOC voted to withdraw recognition. In 2026, the Court of Arbitration for Sport dismissed IBA's appeal. In February 2026, the IOC granted provisional recognition to a new body, World Boxing. In March 2026, boxing was proposed for reinstatement to the Los Angeles 2028 Olympic programme.

Three organisations, three names, one shared question: who holds the right to define clean.

In Asia, where I live and work, the dominant model is event-based testing. A regional promotion hires a collection agency for a few major bouts, pays per sample, publishes a short statement. Event-based testing is the weakest form of testing because it is scheduled. Fighters know the collection date before stepping on the scale. A scheduled test does not measure the body's baseline. It measures the final fortnight.

Three structural layers

The first layer is the money reaching the laboratory. The network of WADA-accredited laboratories is limited globally, and those labs receive samples from state agencies and private organisations alike. Standard procedure requires the sample to be coded before handover. The laboratory receives a number, a sample type, an analysis request. It does not receive a name.

The laboratory does not know the fighter's name. That is why I trust them.

Trusting the laboratory, however, is not the same as trusting the whole chain. The laboratory answers one question: what substance is in this sample, at what concentration. Every question after that — whether the substance is on the prohibited list, whether the athlete holds a valid therapeutic use exemption, whether the concentration crosses a threshold, whether the threshold was crossed by a cold remedy — belongs to a panel appointed by the paying party.

Therapeutic use exemptions are where things start to blur. A fighter with a knee injury treated with growth hormone may obtain a valid exemption. The same hormone, used outside that framework, is a violation. Two medically similar files can end in two opposing legal conclusions, depending on the filing date and the signatory. In my spreadsheet I keep both columns: exemption filing date and sample collection date. The gap between those two dates decides many cases, not the laboratory.

The second layer is data continuity. The athlete biological passport, an approach WADA introduced in 2026, does not hunt for a specific substance. It tracks markers such as the testosterone-to-epitestosterone ratio, haemoglobin and the red blood cell network over time. A single sample says nothing. Three years of data says a great deal.

The third urine sample shows what the first two did not dare to say.

That is why changing the programme administrator matters more than it appears. Every switch of collection agency, analytical provider or retention policy cuts the baseline. Stored samples can be re-analysed years later, but only if they were retained and only if someone pays for that analysis. A new programme holds full discretion over what is stored, for how long, and most importantly, whether anything is re-tested.

In my spreadsheet, I mark red every period in which a testing programme changed administrators. There have been four such instances in twelve years, across four different promotions. In three of the four, the published sample volume for the following year fell against the previous year. None came with a technical explanation. That is an observation, not a conclusion.

I am not saying the 2026 handover was designed to erase data. I am saying no clause in any public document I have read obliges the incoming programme to preserve the length and density of the previous data chain. No such legal obligation exists, because no one holds the authority to impose one.

The third layer is the contract. It is the least discussed layer and the heaviest.

A contract usually runs to one page. A dirty contract has an annex.

A fighter's contract with a promotion determines where disputes are resolved. When accused of an anti-doping rule violation, fighters do not ordinarily go to a civil court. They enter arbitration under a clause they already signed. Legal fees, medical expert fees and independent sample analysis are paid by the fighter. Someone on a twelve-thousand-dollar base purse cannot fund those three items. The result is that most cases end in settlement, not in judgment.

In the Tianhai file I pursued for three years, the recurring lesson was the same: transactions matter more than testimony, and clauses matter more than transactions. The same holds here. A test result is a fact. An arbitration clause is a fate.

The other side of the table

There is an argument I encounter regularly and consider half right. It says the public model has been captured, so private organisations handling their own affairs are not necessarily worse. The evidence sits in WADA's own history. The 2026 McLaren report described a state-run doping system in Russia operating over years, alongside sample swapping at a previously accredited laboratory. Years of argument followed over reinstating or continuing to suspend the Russian anti-doping agency. A system designed to supervise itself demonstrated that it could not.

The wrong half is the comparison. The public model fails at the state level, where sovereignty, budgets and political motive exist. The private model fails at the contract level, where money and clauses exist. Two different failure modes require two different fixes. Importing the public model into a private organisation does not solve the private model's problem. It only changes the signatory.

What both models share is a single assumption: the athlete is the only party with a burden of proof. There is no parallel mechanism on the other side. Laboratories need not publish error rates. Collection agencies need not publish the share of samples rejected for technical faults. Programmes need not publish samples actually collected against samples planned. Those numbers exist, inside internal reports.

A view from the floor

Based on my experience watching fights, there is a parallel between scoring and testing that I have not seen fully analysed. Both rest on an open phrase. In football's VAR, that phrase is clear and obvious error. In combat sports, it is significant strike under the unified rules. There is no arithmetic definition of a significant strike. A leg kick that does not move an opponent can count if a judge reads it as control. A punch that opens a cut without costing position may not count at all.

The subjective space in both systems is wider than viewers assume. And as with VAR, that space is shaped by whoever holds the power to set the definition first. A panel appointed by the paying party can operate with complete transparency, follow every procedure, and still produce an outcome no one on the athlete's side can verify, because the definition sits on their side of the table.

Where the risk sits

Three tiers of risk, ranked by the severity I assign them.

The highest is systemic risk: a testing programme under no obligation to publish internal figures cannot be assessed. A programme that cannot be assessed cannot improve, and cannot be refuted either.

The middle tier is data risk: the biological passport loses value over time if the baseline is cut. This needs no conspiracy. It occurs as the natural consequence of changing a service provider.

The lowest tier, and the most expensive in human terms, is individual legal risk: fighters without the money to defend themselves inside a system they signed into before they had a career.

On betting

There is another thread I track in parallel. Quarterly reports from the International Betting Integrity Association consistently place combat sports and esports among the categories with above-average suspicious odds alerts. For esports, the explanation offered is a regulatory gap. For combat sports, I would argue the cause is similar but more specific: when a bout outcome depends on a phrase with no arithmetic definition, the betting market has a grey zone to operate in, and that grey zone breaks no rule.

A strong anti-doping programme will not fix that. A weak one makes it worse, because it devalues physical data — the only thing that can be cross-referenced against what happens on the floor.

Takeaway

I do not expect a new organisation. The next meaningful reform is not another supervisory body but two technical requirements that could be written into contracts this season: periodic publication of samples collected against samples planned, and a mandatory legal fund for accused athletes. Neither requires trust. Both are measurable. And both can be cross-checked by anyone with a spreadsheet and enough patience.

If a programme will not publish its own sample counts, the only remaining credible number is the one inside the contract.