The survey queried the entire population of known drug courts in the United States. Survey responses were received from 257 of the 340 drug courts that were operational as of mid-1999 (response rate of 75.59 percent). The survey found that generally many drug court processes are not automated, and the automated support being provided to drug courts by existing computer systems is inadequate, such that drug court operations are hampered in many ways. They cannot take advantage of productivity gains and success-sharing that computer automation offers, and they cannot serve larger numbers of participants unless they are more fully automated. A strong correlation was found between automation and the time it takes for the drug court judge to receive failed drug test results. A lack of funding was reported to be the primary reason drug courts have not acquired additional automation to improve court administration. The difficulty of linking to other systems was another barrier to automation. Less than 15 percent of the surveyed courts reported that they have completed the necessary automation required to produce reports needed for overall program evaluation. The drug courts overwhelmingly expressed a desire for additional education and training to deal with evaluation and management information systems. Among the recommendations are that specialized educational programs be developed to address high-priority issues; that technical assistance address areas identified as needing special attention; that a software package tailored for drug courts be developed; and that a research agenda be developed for drug-court best practices. 7 figures, 7 tables, and appended survey questionnaire and write-in comments