The STAR model takes a systems approach to estimate the flow from producer countries through the transit zones, across borders, and throughout the United States. The model breaks cocaine movement down into a series of stages based on the cultivation, production, transportation, and marketing of the product. The model estimates cocaine availability at each stage by triangulating between three dynamic existing processes. The first process estimates coca cultivation based on overhead imagery. The second estimates cocaine leaving South America based on tabulation of movement events. The third process estimates United States consumption based on prevalence estimates and trends in cocaine price and purity. Data sources include the Federal Drug Seizure system, the National Household Survey on Drug Abuse, the Consolidated Counterdrug Database, the Arrestee Drug Abuse Monitoring Program, coca cultivation data, and the Drug Enforcement Administration’s System to Retrieve Information on Drug Evidence. The estimates reveal that as expected, the potential production estimates exceed all other estimates every year and that the estimate of domestic consumption has been stable at 500-600 metric tons over the past 4 years from 1996 through 1999. The analysis concludes that the STAR model provides a means of incorporating differing data within a cohesive structure and that a consumption-based estimate of cocaine availability provides a consistent approach for integrating the various data set. Figures, tables, footnotes, appended tables and background information, and 19 references