These data were gathered to assess whether drug users are greater risks than nonusers for rearrest or failure to appear for scheduled court appearances while on pretrial release. The data also evaluate the relative effectiveness of periodic surveillance through urinalysis, traditional narcotic treatment, or neither in reducing rearrest and failure to appear during the pretrial period. The collection provides information on arrestees who both tested positive for drugs and were released on recognizance as well as those arrestees who tested negative but were not released on recognizance. Drugs tested for include heroin, cocaine, PCP, methadone, and amphetamines. Arrestees who were released were randomly assigned to one of three groups: weekly urine testing, referral to drug treatment, or a control condition. The data offer information on the offender's background, family and employment status, probation and parole status, pending charges, and prior convictions. Other variables include date of arrest, charge, initial release, decision, date of disposition, type of final disposition, number of subsequent arrests before trial, and number of bench warrants issued. Results of urine tests at arrest are available for about 65 percent of the total sample. For those in the experimental surveillance group, summary urine test results from the periodic testing program are available. There is no measure of treatment for the drug treatment or control groups.