SUMMARYWe use data from the British Crime Survey (BCS) to analyse the relationship between illicit drug use and labour market outcomes for a sample of men and women aged 16 to 25. In using this data we highlight a serious design flaw in the BCS questionnaire structure that presents a serious barrier to statistical modelling of drug use at the individual level. We propose a simple way of overcoming this problem and proceed to estimate a model of occupational attainment jointly determined with unemployment and current drug use, conditional on past drug use. Separating the commonly abused drugs into a "hard" and "soft" category, we find that past hard drug use has a significant positive association with current unemployment, but find no significant association between past hard or soft drug use and occupational attainment. We also find no significant association between current drug use and attainment, although we observe that current drug use is associated with current unemployment. We suggest that previous research has tended to find a positive relationship between drug use and wages because of a failure to take into account current labour market status.KEY WORDS -Illicit drug use, unemployment, occupational attainment.