It further indicates -- through a meta-analysis that pooled comparisons of reported drug use by 16- 24-year-olds -- that added value can be obtained from the existing data sets; for example, previously unrecognized and substantial increases in reported use of amphetamine or ecstasy were found by pooling trends across British general population surveys. These increases of one-third occurred in the mid-1990's and were revealed because secondary analysis led to increased precision. Even with this enhanced precision, however, current monitoring strategies are still underpowered. Besides recommending enlarged general population surveys, this paper recommends that for the monitoring of youths' injecting or use of heroin, alternative surveys should be commissioned to target higher risk sentinel groups; such groups might include youth living in deprived areas or young offenders. Survey sizes must be larger, and survey sites must be appropriately selected to ensure that regional variation in illegal drug use is properly addressed. 6 tables and 18 references