AIMS: To describe and evaluate trends in the use of stimulant drugs over a 5-year period using an under-studied data collection method. DESIGN: Repeated-measures cross-sectional survey. SETTING AND PARTICIPANTS: Annual magazine-based survey targeting people who use stimulant drugs in dance contexts. MEASUREMENTS: Life-time use prevalence (ever used), age of first use, current use prevalence (any use within the last month) and extent of use within the last month (number of days used) for a range of stimulant drugs. Additional measures of quantity of ecstasy used were also collected. FINDINGS: Trends in life-time and current prevalence over time have been detected and comparisons made between different stimulant drugs. Evidence is obtained of broad stability in patterns of stimulant use in respect of age of first use and frequency of use among ongoing users. Despite an apparent reduction in the current prevalence of ecstasy use, the proportion of heavy users (usually >4 pills per session) has more than doubled between 1999 and 2003. CONCLUSIONS: This purposively sampled population study has yielded time trend data broadly consistent with other indicators, where they exist, and also has demonstrable potential to identify new drug trends. Further comparisons of purposive samples and randomly formed samples are needed.