Drug-use incidence, as perhaps measured by the fraction of successive youth cohorts entering into drug use, is a prevalence derivative that could be a major influence on drug-control budgets, since this indicates the extent and nature of the problem. Other prevalence-related figures (consumption, expenditures by drug users, and drug production) are suitable as guides for decisions about resource allocation among programs; for example, consumption estimates can provide a measure of the effectiveness of interdiction by indicating what share of shipments are being seized. This might help guide decisions about allocations among various kinds of enforcement. That prevalence estimation has not been used in such policymaking in American drug strategy is due to both the limited capacities and credibility of prevalence estimates and the moralistic nature of the policy process associated with illicit drug use. The available numbers are either developed systematically from data sources that have low credibility (self-report) or are developed less systematically from sources that are not well understood. For prevalence of drug use, both trends and absolute levels can be estimated with tolerable accuracy, but the policy process is weakly linked to such estimates. Prevalence estimation could have its most important use at the local level, but the per capita cost of prevalence estimation is much higher for metropolitan areas than for the Nation, so it is not likely that it will be used extensively in the policy process. The expansion of the National Household Survey of Drug Abuse may allow some major States and some metropolitan areas to develop estimates from that survey. 23 notes and 36 references