All 35,297 cases of drug-related arrests reported for Quebec in 1997, 1998, and 1999 were extracted for the study. A capture-recapture methodology was used to analyze the data. This distinguishing feature of such a technology is that the odds of arrest are estimated by defining a population of "susceptibles," namely, a population of individuals involved in a particular area of activity (burglars, prostitutes, car thieves, heroin dealers) who have yet to be arrested but who nonetheless have the same characteristics as arrested offenders. The higher the odds of rearrest in a given distribution of those arrested for a given offense, the smaller the pool of susceptibles yet to be arrested. Odds of arrest estimated by capture-recapture techniques (how many arrests for how many susceptibles) provide a crude indication, at the aggregate level, of the likelihood that the pool of motivated and ready-to-act offenders participating in a given criminal activity will be arrested for a related offense. Findings showed that the risks of being arrested were much higher for sellers than for consumers and that this gap widened for the more harmful drugs. Findings also indicated, however, that vulnerability to arrest was significantly higher for marijuana users than for users of other drugs, and that dealers in the smaller but more harmful drug markets (crack and heroin) experienced lower aggregate risks of arrest than cocaine or marijuana dealers. Suggestions are offered for future research. 3 tables, 5 notes, and 49 references