This report compares the research methods, provider and patient characteristics, and outcome results from four large-scale followup studies of drug treatment during the 1990s: (1) the California Drug and Alcohol Treatment Assessment (CALDATA); (2) Services Research Outcomes Study (SROS); (3) National Treatment Improvement Evaluation Study (NTIES); and Drug Abuse Treatment Outcomes Study (DATOS). The first two studies were, respectively, state and national level stratified random probability samples of provider and patients, and the latter two were prospective repeated interview designs focusing on urban networks. All studies received multistage response rates of about 60%, except DATOS, which achieved about 40%. From before treatment to after treatment, each study revealed across-the-board positive changes in drug use, crime, and health status. Clients in SROS shows consistently fewer changes than others on comparable variables, something that may result from its longer reference periods and the lower precision of the variables selected to measure long-term change. Researchers should further the potential of these large data sets to capture differences in client populations, organization of services, and clinical practices that may explain differential outcomes and costs. An appendix describes the NTIES. (Contains 16 exhibits and 9 references.) (Author/SLD)