Three principal issues are addressed: the factors that affect processing times, by how much, and under what conditions. The proposition underlying the research is that everything directly affecting processing time comes under one or more of five headings: the burden of work in relation to resources, the complexity of the case, events in the life of the case, participant incentives, and structural and administrative facilitation. Three courts, each of which saw major structural aad administrative changes during the study period, are examined, and the effects of data are estimated. A model of case processing time is developed, and its parameters are estimated from data of one State criminal court. The model is then extended to accommodate additional court-level variations, and estimations for two more cities are provided. Results of such variables as number of defendants, disposition type, pretrial motions, failure to appear, continuances, and psychiatric evaluation and treatment are compared. Finally, a distributed lag model is proposed and estimated. The model expresses a court's pending caseload as a multiplicative function of the number of new filings and the mean processing time for each of a number of preceding months. Twenty-four references and three tables are provided.