Cycling doping history explained, era by era

Cycling's doping history is usually told as a list of scandals. Told that way it is a parade of names and bans that explains nothing. Told as a sequence of problems and the detection methods each one provoked, it becomes the clearest case study in sport of how an anti-doping system actually gets built, which is almost always after the damage rather than before it.

The history turns on five dates. A death on a mountain in 1967. A car stopped at a border in 1998. A Spanish police investigation in 2006. A new kind of test in 2008. And a reasoned decision in 2012 that removed seven Tour de France titles and put nobody in their place.

Each of those moments changed what the sport was trying to detect, and each of them arrived because the previous method had stopped working. Here is the sequence.

• 1967Tom Simpson dies on Mont Ventoux
• 1998The Festina affair breaks during the Tour
• 1999The World Anti-Doping Agency is established
• 2006Operación Puerto becomes public
• 2008The athlete biological passport is introduced
• 2012Seven Tour de France titles are stripped

Each date marks a change in what the sport was trying to detect or in who was doing the detecting, rather than simply a scandal.

Why cycling became the test case

Cycling did not invent doping and is not the only sport with a history of it. It became the sport where anti-doping was built because of four features that combine badly.

The physiology is unusually responsive. Endurance performance over several hours is limited substantially by oxygen delivery, and oxygen delivery is one of the few things in human physiology that can be raised dramatically by pharmaceutical means. A drug that increases red cell mass does not make a rider marginally better; it moves them into a different competitive bracket. In sports where the limiting factor is skill, technique or short-duration power, the available gains are smaller.

The events are extreme. A three-week stage race asks for sustained effort far beyond what recovery would normally permit, repeated daily. The incentive to find pharmaceutical help with recovery is structural rather than moral, and it predates any modern drug by the better part of a century.

The economics are precarious. Professional cycling runs on sponsorship rather than broadcast rights or ticket sales, team budgets are small by the standards of major sport, and contracts are short. A rider who does not deliver results does not get renewed, and the consequences arrive within months.

The sport was institutionally close to its own medicine. Team doctors and soigneurs were part of the structure, which meant a rider seeking help had somewhere to seek it and a team seeking performance had a mechanism. That proximity is what turned individual choices into organised programmes.

None of this excuses anything. It explains why the problem in cycling reached a scale that forced the creation of institutions that now govern every sport, and why the detection methods used across the Olympic movement were largely developed in response to what was happening in bike racing.

The amphetamine era, and the death that forced the first rules

The first era is stimulants: amphetamines and related compounds, used to suppress fatigue and pain and to allow a rider to keep going past the point at which they would otherwise have stopped. This was not a secret practice in any meaningful sense. It was widespread, it was discussed, and in the absence of any rule against it there was nothing to conceal.

The accounting changed on Mont Ventoux in 1967, when Tom Simpson died during the Tour de France. Amphetamines were a factor, alongside extreme heat and dehydration, and the combination was precisely the one stimulant use produces: a rider chemically prevented from registering how close to collapse he was, in conditions that required him to register it.

A death on television during the sport's biggest race is a different kind of event from a rumour about a peloton, and it removed the argument that this was a private matter between riders and their own bodies. The first serious controls followed, and the riders' initial response was to protest against them, which is a detail worth keeping rather than skipping. The protest was not simply self-interest. Testing was new, the procedures were crude, and a rider's career could be ended by a process nobody trusted. The relationship between athletes and anti-doping has been shaped by that opening exchange ever since.

What the first controls could do was limited. They were in-competition only, they looked for a short list of substances, and the detection windows were narrow. The practical effect was to make stimulant use require timing rather than to stop it.

Blood, before anybody could test for it

The second era changed the target from the nervous system to the blood, and it did so with a method that no chemical test could ever find.

Blood doping in its original form is mechanical rather than pharmaceutical. Remove blood, let the body replace it, reinfuse the stored blood before competition, and the rider is competing with more red cells than their physiology would otherwise supply. There is no foreign substance involved. The blood is the rider's own.

That is the important point and the reason this era matters. The sport's entire detection apparatus was built on the assumption that doping means introducing something that should not be there, and here was a method that introduced nothing. The practice was eventually prohibited, which is a separate thing from being detectable: prohibition without detection shifts a practice from open to concealed and does little else.

A test capable of identifying blood from another person arrived later, and its effect was precise and limited. It made transfusing someone else's blood dangerous and left reinfusing the rider's own blood as difficult to detect as it had always been. Each detection advance in this history produced a narrowing rather than an ending, and this is the cleanest early example.

EPO changed the arithmetic of endurance

The third era is the one that did the most damage, because it combined a very large performance effect with a very long period of undetectability.

Erythropoietin is a hormone the body produces to regulate red cell production. A synthetic version developed for patients with kidney disease and anaemia does the same job, and administered to a healthy endurance athlete it raises the oxygen-carrying capacity of the blood substantially. It achieves what a transfusion achieves without the logistics of storing blood, and it is administered by injection in quantities that fit in a pocket.

Two consequences followed, and the second is less discussed than the first.

The performance effect was large enough to reorder the sport. In an event where sustainable power over hours is the binding constraint, raising oxygen delivery raises that power directly, and the figures that decide a mountain stage are precisely the figures that respond. The relationship between sustainable power, body mass and climbing speed set out in the piece on watts per kilogram is the mechanism: anything that lifts the numerator lifts climbing speed proportionally, and EPO lifts the numerator.

The second consequence is that the drug was difficult to detect because the body makes it too. Finding an administered version of a substance the athlete produces naturally is a far harder analytical problem than finding a foreign compound, and for a period there was no validated way to do it at all. A direct urine test for synthetic EPO became available at the start of the 2000s. Before that, the sport knew the drug was in use and could not prove it in any individual case.

That gap, between knowing and proving, is where the worst of this history happened.

The 50 per cent rule, a health check standing in for a test

Faced with a drug it could not detect, the governing body did something revealing. In the 1990s it set a ceiling on haematocrit, the proportion of blood volume made up of red cells, at 50 per cent. A rider measured above it was declared unfit to start on health grounds and stood down, rather than being sanctioned for doping.

The health justification was genuine. Very high haematocrit thickens the blood and raises the risk of clotting, and that risk is not theoretical in an athlete who is dehydrated, exerting hard and whose heart rate falls low in sleep. The sport had real reason to be worried about what was happening to riders' blood.

But the rule's actual function was substitution. Unable to detect the cause, the sport capped the effect, and in doing so it published a number that told anyone using the drug exactly how much they could use. A ceiling expressed as a measurable value is, from the point of view of someone doping carefully, a target. The rule converted an open-ended risk into a managed one.

This is the single most instructive episode in the whole history, because it shows what happens when a sport regulates a proxy rather than the thing it cares about. The 50 per cent limit was honest about its own inadequacy, it probably did prevent some genuinely dangerous practice, and it simultaneously made careful doping more viable than careless doping. Every subsequent design decision in anti-doping has been shaped by the memory of it.

1998: the Festina affair and the end of plausible deniability

The 1998 Tour de France is the hinge of this history, and what made it decisive was not a test result.

A team soigneur was intercepted carrying a large quantity of doping products. Police raids followed, a team was expelled, riders were questioned, and the race continued through protests and withdrawals while a criminal investigation ran alongside it. The sport's biggest event spent three weeks being investigated in public.

Three things changed permanently.

The evidence changed hands. A police investigation can search premises, seize records, compel testimony and follow money. A laboratory can analyse a urine sample. Those are not comparable powers, and the Festina affair demonstrated that the most effective anti-doping instrument available was not a test at all. Every major case since has leaned on investigative evidence.

The scale became undeniable. What was found was not a rider's personal supply. It was an organised, team-level system with records, which meant the problem could no longer be framed as individual riders making individual choices. Teams, doctors and management were implicated structurally.

The sport lost the ability to self-police credibly. An institution that had been asserting the problem was under control was shown, during its own showpiece, not to have known the scale of it. That loss of credibility is what made the next step possible.

WADA, and taking the problem out of the sport's hands

The World Anti-Doping Agency was established in 1999, directly in the wake of that year.

The structural insight behind it was simple and had been obvious for a long time without being actionable: a sport cannot credibly investigate itself, because it has a commercial interest in the answer. A governing body that uncovers a large doping programme damages its own product, its sponsors and its broadcast value. Even acting in complete good faith, it is the wrong body to be making those judgements.

WADA's answer was to move the rules, the prohibited list, the laboratory accreditation and the appeal route outside any individual sport, and to bind sports to a common Code by contract rather than by law. The architecture of that system, and the reasons it works the way it does, is set out in the piece on how the World Anti-Doping Code operates.

What matters for this history is the direction of the change. Before 1999, anti-doping in cycling was something cycling did to itself. After it, the standards were set externally, the laboratories were accredited externally, and a decision the sport reached could be appealed by a body with no stake in cycling's commercial health. That is the single largest governance change in the sequence, and it came out of cycling's worst year.

Operación Puerto, and what police can see that laboratories cannot

In 2006 a Spanish investigation became public, and it made the Festina lesson explicit.

What the investigation found was a medical operation: stored blood, documentation, a client list identified by code. No sample analysis had produced any of it, and no sample analysis could have. The evidence was physical and documentary, obtained under powers no sporting body possesses.

Two durable consequences followed.

The first is that cycling's anti-doping effort became permanently intelligence-led. Testing continued and expanded, and alongside it the sport and the anti-doping organisations began treating information, patterns, associations and documentary evidence as primary rather than supplementary. Several of the rule violations in the modern Code, including possession, trafficking, administration and complicity, are designed to be provable without any positive test, and they exist because cases like this one showed they had to.

The second is messier. Investigations of this kind generate evidence about many people, held by prosecutors operating under criminal procedure in a particular jurisdiction, with its own rules about disclosure, destruction and the rights of people who have not been charged. The result was years of legal argument about who could see what, and a widespread public impression that the sport had declined to pursue what it had been handed. Some of that impression was unfair and some of it was not. What is not in dispute is that a sporting body trying to use criminal evidence is operating at the mercy of a legal system designed for a different purpose.

2012: seven titles and no replacement

In 2012 the United States Anti-Doping Agency published its reasoned decision, and Lance Armstrong was stripped of seven Tour de France titles.

The case is the clearest demonstration in sport of a point that still gets missed. It did not rest on a positive test. It rested on witness testimony, documentary evidence, financial records and the pattern they formed together, assembled by an agency with the resources to pursue a case over a long period against a well-funded defence.

The most informative detail is what happened to the results afterwards. The race lists no winner for those years. Nobody was promoted.

That decision is worth sitting with, because it is an admission the sport made about itself. Promoting the second-placed riders would normally be the automatic remedy. It was not applied, because the governing bodies concluded that doing so would not produce a credible result. A blank space in a palmares is a governing body saying, in the only language a record book has, that it does not believe any of the available answers.

The years remain empty. Whatever else is arguable about how cycling handled its history, that particular piece of accounting is honest in a way that sport rarely manages.

1. 1StimulantsAmphetamines and related compounds used to suppress fatigue and pain. Widespread before any rule prohibited them, and the practice that produced the sport's first controls after a death on Mont Ventoux in 1967.
2. 2Mechanical blood dopingRemoving, storing and reinfusing a rider's own blood. It introduced no foreign substance at all, which defeated the entire logic of a system built to find things that should not be present.
3. 3Synthetic EPOA pharmaceutical version of a hormone the body already produces, raising oxygen-carrying capacity substantially. Difficult to detect precisely because the natural version is always present, and in use for a period during which no validated test existed.
4. 4The haematocrit ceilingThe sport's response to being unable to detect EPO: a 50 per cent limit on red cell proportion, enforced as a health check rather than a sanction. It capped the effect and, by publishing a number, also defined a safe dose.
5. 5Designer steroids and growth factorsCompounds synthesised specifically to fall outside the known analytical targets, and peptide hormones the body produces naturally. Both attack the same weakness: a test can only look for what it has been designed to find.
6. 6Investigative evidencePolice raids, seized records, testimony and financial trails. Not a detection method at all, and the single most effective instrument in the history, because it can see organisation rather than only substances.
7. 7The biological passportMonitoring a rider's own blood and steroid markers over time and flagging patterns normal physiology would not produce. The question moves from what is in this sample to whether this body is behaving as it should.
8. 8MicrodosingDoses small enough to clear quickly and to keep markers inside a rider's established ranges. The practice the passport was built to expose and the one it is least certain to catch, which is where the arms race currently sits.

Each era is defined by what the previous detection method could not see. The sequence runs forward in time, and each entry made the next detection advance necessary.

The biological passport: from the substance to the fingerprint

The Union Cycliste Internationale introduced the athlete biological passport in 2008, and it is the most important technical change in this history.

Every method before it asked the same question: is there a prohibited substance in this sample? That question has a structural weakness, which is that it can only be answered for substances the laboratory is looking for, within the window during which they remain detectable. A programme designed around detection windows beats it, and a substance the body produces naturally beats it twice.

The passport asks a different question. It builds a profile of the rider's own blood and steroid markers from repeated samples over time, establishes what is normal for that individual, and flags departures that doping would produce and ordinary physiology would not. No substance has to be identified. The evidence is the pattern.

Three things follow from that design, and all three changed behaviour.

Avoiding a positive test stopped being the same as avoiding detection. A rider could clear every sample and still accumulate a profile that could not be explained.

Every sample became valuable. Under the old logic, a negative test was a null result. Under the passport, a negative sample is a data point that contributes to a profile, which means the samples that find nothing are doing work.

The deterrent became permanent rather than episodic. A detection window closes. A profile does not.

The passport is also slower, more expensive and harder to argue at a hearing than a laboratory finding, because an inference from a pattern is contestable in ways that an identified molecule is not. The machinery of how a flagged profile becomes a case, and the expert review it passes through, is covered in the piece on how drug testing works.

Microdosing, and the arms race the passport changed rather than ended

The honest assessment of the passport is that it moved the frontier rather than closing it, and the sport's own published concerns say so.

If a profile flags departures from a rider's established ranges, the counter-strategy is to dope in quantities small enough to stay inside those ranges. Smaller doses, administered more often, timed around likely testing. The performance benefit is smaller than a large programme would deliver, and it may be large enough to matter in a sport decided by small margins.

Whether the passport reliably detects this is a genuine open question rather than a rhetorical one. The argument that it does is that any dose worth taking has to change something measurable, or it would not be worth taking. The argument that it does not is that the statistical model needs a departure large enough to exceed a rider's natural variation, and a carefully managed programme can operate below that threshold.

Both arguments are reasonable, neither has been settled publicly, and the sport is in the position of running a system whose effectiveness it cannot fully characterise. That is an uncomfortable position and it is more honest than the alternative of declaring the problem solved.

What the sport actually changed

Beyond testing, cycling made structural changes that are easy to overlook and probably did more than any single detection method.

The no-needles policy. Injections without documented medical justification were prohibited, which removes a practice that had been normalised as recovery medicine and was the delivery route for a great deal else. It is a rule about behaviour rather than substances, and it is enforceable by inspection.

Separating teams from their own medicine. Team doctors having both a duty of care and a stake in performance was the structural flaw underneath the organised era. Independent medical oversight, restrictions on who can treat riders, and reporting obligations all attack it.

The whereabouts system. Riders in a registered testing pool must say where they will be, so that out-of-competition testing can be genuinely unannounced. This is the rule athletes dislike most and the one without which out-of-competition testing does not function, because a test that can be anticipated is a test that can be planned around.

Long-term sample storage. Samples can be stored and reanalysed later as methods improve, and the Code permits a case to be brought within ten years. This changes the risk calculation fundamentally: a rider doping with an undetectable method today is betting that it will still be undetectable in a decade.

Team-level accountability. Licensing conditions, internal programmes and the willingness of sponsors to leave have made a doping case an existential risk for a team rather than a problem for one rider. A structure that punishes only the individual leaves the organisation intact to do it again, and the recruitment and contracting pressures described in the piece on grand tour team selection now operate against a background where a team's licence is at stake.

• 50Haematocrit ceiling set in the 1990s, per cent
• 10Years within which a case may be brought under the Code
• 0Tour de France winners listed for the stripped years

Each figure is a published decision rather than an estimate, and each one changed behaviour in a different direction.

The cost borne by the riders who did not dope

Any account of this period that deals only with institutions and detection methods leaves out the people it was worst for, and they were not the ones who were caught.

A rider competing clean during the undetectable era faced a choice with no good option. Compete as they were and lose to riders operating at a level their physiology could not reach. Join the practice. Or leave the sport. The first produced a career spent being told they lacked talent. The third produced no career at all. Several riders have described exactly that calculation in public, and it is the most damaging consequence of a long gap between knowing and proving: the people who behaved correctly paid for it, in results, in contracts and in reputation, and no governing body can give any of that back.

The point has a practical edge rather than only a moral one. A doping era does not simply add illegitimate results at the top; it reorders the entire field, because every placing is a comparison. A rider who finished fortieth in a race where the first twenty were doping was not the fortieth best rider there, and no reallocation of a trophy repairs that. It is why the empty years in the Tour's record are a more honest remedy than promotion, and why they are still an inadequate one.

It also explains why the sport's internal culture changed more slowly than its rules. Riders who came through that period carry the knowledge that the system did not protect them, and riders who came after inherit a sport where suspicion attaches to any exceptional performance by default. A clean rider today producing a genuinely outstanding climb will be questioned, publicly, on the basis of a history they had nothing to do with. That is an unfair cost and it is a direct consequence of the years when the sport could not tell the difference, and it will keep being paid for as long as the only available response to an extraordinary performance is to wonder.

The open questions, stated honestly

A history like this invites a tidy conclusion and does not support one. Four questions remain genuinely unresolved.

Can the absence of doping ever be demonstrated? No, and this is a logical point rather than a cynical one. A testing system can establish that it found nothing. It cannot establish that there was nothing to find. Anyone claiming a clean peloton is making a claim the evidence cannot support, and anyone claiming the opposite is doing the same.

Is retrospective testing fair? Reanalysing a stored sample with a method that did not exist when it was given catches people who could not have known they were detectable. That is exactly the deterrent effect the ten-year window is designed to produce, and it also means a rider is judged by a standard that arrived after the act. Both of those statements are true, and the sport has chosen the deterrent.

Does the passport catch careful doping? Covered above, and unresolved. The sport's position is that the risk is high enough to deter. The counter-position is that the threshold is knowable and therefore manageable.

Is the burden distributed fairly? Testing intensity depends on the anti-doping organisation responsible for a rider, and those organisations differ enormously in budget and capability. Two riders on the same start line may have faced very different levels of scrutiny over the preceding year, and neither chose it.

How to read a cycling doping story

Four questions make almost any report on this subject more legible.

Is this a test result or an investigation? A laboratory finding and an investigative case are different kinds of evidence with different strengths. Nearly every large case in this history came from investigation, and nearly every headline implies a failed test.

Is it a substance case or a pattern case? A finding identifies a molecule. A passport case infers from a profile and is argued by experts. The second is harder to report accurately and more easily misdescribed as the first.

Which era's method is being discussed? A story about detection windows is a story about the direct-testing model. A story about reference ranges is a story about the passport. Conflating them produces nonsense.

What stage has the process reached? A reported finding is not a sanction, and a first decision is not final. Cases in this sport have been reported as concluded at three or four separate points, and the history is littered with headlines that outran the procedure.

Read that way, cycling's doping history stops being a parade of disgraces and becomes something more useful: a record of how a sport discovered, repeatedly and expensively, that it could not detect what it was trying to detect, and of the institutions that exist now because of it. The rest of the cycling explainers here take the same approach, beginning with the mechanism before asking what it was built to solve.