UK and US literature indicate that cocaine and crack users experience multiple problems and poor treatment outcomes (Gossop et al., 2002 , 2003 ). Using data collected as part of a Scottish national evaluation of drug treatment effectiveness, this paper: (i) provides information on the nature and extent of recent cocaine and crack use among 585 new drug treatment clients; and (ii) investigates whether those who had recently used cocaine and/or crack comprised distinct treatment subpopulations. Data were analysed using chi-square tests of association (or Fisher's exact tests), ANOVA and stepwise multivariate logistic regression. Although more respondents than anticipated reported cocaine and/or crack use, consumption was generally infrequent and not considered problematic. Recent cocaine and/or crack users were more likely than other respondents to have taken non-prescribed benzodiazepines in the last 90 days; be seeking residential treatment; have recently committed acquisitive crime; have been robbed, attacked or assaulted in the last six months; have a spouse or partner; and have a longer history of problematic drug use. In contrast, few differences between those who had recently used "cocaine only", "crack only", and "both cocaine and crack" were found. It is concluded that drug agency staff in Scotland must be alert to, and have the resources and training to deal with, the complex needs of the increasing numbers of cocaine and crack users approaching their services.