Anti-doping is usually discussed as a contest between chemists. In practice most cases turn on procedure, because a result is only usable if every step from collection to analysis can be documented.
Collection is designed to remove doubt
An athlete selected for testing is accompanied by an officer from the moment of notification until the sample is sealed, so there is no unobserved interval.
The athlete chooses the collection vessel and the sealed kit, splits the sample into two bottles, and seals them personally.
Every step is recorded on a form the athlete signs, including any objection, and that document becomes part of the evidence if a result is later contested.
Two bottles exist for a reason
The sample is divided into an A and a B portion, and only the A portion is analysed initially.
If the A result is adverse, the athlete may request that the B portion be opened and analysed, with the right to attend or send a representative.
A B result that does not confirm the A result ends the case, which is why the split is the central procedural protection in the system.
Testing happens away from competition too
Many substances are taken during training rather than at events, and their traces have gone by the time an athlete competes.
Out-of-competition testing addresses this, which requires athletes in certain pools to file information about where they can be found each day.
Missed tests and filing failures carry consequences of their own, and a pattern of them can result in a sanction without any positive analysis.
The biological passport looks for change
Rather than detecting a substance, a passport tracks an athlete's own blood and steroid measurements over years.
Values that shift beyond what normal physiology and training explain can indicate manipulation even when no prohibited substance is found.
Because the comparison is against the athlete's own baseline, the approach catches methods designed to leave no direct trace, and it is assessed by expert panels rather than by a single threshold.
Samples are stored for later analysis
Samples are retained for a period of years, and rules allow them to be reanalysed as detection methods improve.
Results have been reversed long after events on this basis, which is the main deterrent against substances that current tests cannot identify.
It also means a clean result at the time is provisional in a way that is rarely reported, since the sample that produced it may still be sitting in a freezer.