Testing was voluntary, and consent rates in both samples were over 90 percent. In both samples, arrestees charged with drug possession or sale were underrepresented, while felony arrestees were overrepresented. Data for 1986 indicate cocaine use in 83 percent of arrestees, marijuana use in a little over 25 percent, opiates in 20 percent, and methadone and PCP in a relatively small percent. A comparison with 1984 data reveals that cocaine use increased considerably (from 42 percent), opiate and methadone use remained relatively stable, while PCP use declined. In general, cocaine use was most prevalent in young arrestees (16 to 20) and in those over 35. Comparisons for 1984 and 1986 marijuana use could not be made because it was not among drugs tested in the first sample. Brief interviews conducted in 1986 indicate that arrestees underreported their recent illicit drug use even in confidential research interviews. Finally, while only a small percent of arrestees were in alcohol or drug treatment, 25 percent indicated a desire for some treatment services. Results underscore the utility of a national drug crime forecasting system in identifying trends in drug use that can provide a basis for planning and resource allocation. 5 tables and 1 figure.