The study used information from urine samples collected from 2,606 persons who were arrested between April and October 1984. The samples were tested for heroin, cocaine, phencyclidine, and methadone. The analysis considered two measures of pretrial misconduct: whether a defendant failed to appear for a scheduled court date and whether a defendant was rearrested before case disposition. Censored probit models were used to estimate the statistical association between drug test results and pretrial misconduct. Results showed that drug test results are significantly associated with pretrial misconduct over and above the information that is typically available to judges at the time they make release decisions. Findings are consistent with those from research in Washington, D.C. Drug test results may be a useful source of information for a judge to consider along with other information when making pretrial release decisions, but several legal and ethical issues must be addressed. Tables, footnotes, and 19 references.