Aims: Dexamphetamine substitution is a widely practised, yet under-researched and controversial treatment for amphetamine abusers. This study aimed to evaluate the usefulness of substitute prescribing, to both oral and intravenous users, and to find out which factors predict doing well in treatment. In the absence of more rigorous controlled trials, it was hoped that this study would help make some inroads into what is a hitherto unexplored area. Design: The standardized records of 220 users receiving dexamphetamine prescriptions were examined retrospectively. Cross-sectional socio-demographic data, and longitudinal outcome data were obtained for 148 of them. Setting: The amphetamine users had all attended and received treatment by Cornwall Community Drug Team, in the far South-West of England, during the period 1992-96. Findings: Oral and intravenous users had remarkably similar outcomes, with intravenous users making more overall gains in treatment. Over half the injectors stopped injecting, and more than a third within 2 months of coming into treatment. Variables predicting a good outcome differed between oral and intravenous users; although for both groups being female was associated with a slower change in drug-use behaviours, but a longer period in treatment. Conclusions: Dexamphetamine prescribing appears to be reasonably safe, and is associated with improvements in drug-use. Randomized trials are warranted to determine the specific efficacy of the treatment.