The belief that drug addiction is the motor behind much property crime has been a key part of recent UK drug policy. Review of the literature shows that the empirical evidence does not support this view. It also demonstrates that the fundamental weakness of existing research is its failure to address the question of causality properly, relying mainly on uni-directional mechanistic ‘cause-and-effect’ models. It is suggested that the drug–crime link needs to be conceived in the context of the interaction and inter-relation of a range of factors operating at different levels and as part of a set of complex processes. The implications for policy are that a reliance on improving access to treatment is likely to be of limited effectiveness in reducing drug-related crime and that an emphasis on tackling social exclusion may be more fruitful. Future research needs to explore further the notion of causality outlined in this paper. It must also examine emerging patterns of polydrug use and look at minority ethnic groups and women.