We provide a critical discussion of the concept drug-related crime and review methods for estimating its volume, emphasising the importance of an appropriately defined counterfactual. We then construct new estimates for England and Wales in 2003-6, combining data from the Arrestee Survey and Offending Crime and Justice Survey to ensure adequate coverage of prolific offenders/drug users and non-household residents, who are under-represented in household surveys. We find, first, that the volume of drug-induced acquisitive crime linked to heroin use is high, but there is no significant evidence of violent crime linked directly to heroin use. Second, we find no evidence at all of any drug-induced crime committed by people who use cannabis (but not heroin or cocaine). Third, we find evidence that supplying cannabis leads to a small volume of crime. The mechanisms linking cannabis supply to criminal activity merit further investigation.