To better understand the role of treatment over time and the factors that extend its benefits are major concerns of the substance abuse field. This is a competing continuation of a longitudinal study of 1204 managed care (Kaiser Permanente) substance abuse treatment patients who were first interviewed in 1994 in a randomized controlled comparison of day hospital and traditional outpatient programs. We propose to collect 9- and 11-year data and use growth curve models and other analytic tools to examine factors influencing 11-year trajectories of substance use, readmission, medical outcomes, utilization, and cost. Survey data (with drug testing) have been collected at baseline, during treatment, 6 and 12 months, and 5 and 7 years. The health plan documents all substance abuse treatment visits and services, and all medical and psychiatric diagnoses, services received, and costs. We also include a sample of Kaiser members matched with our sample on age, gender, and length of enrollment. These data sources combined afford an opportunity to examine the medical, psychiatric, and substance abuse outcomes, utilization and costs of a well-characterized sample of commercially insured substance abuse treatment patients. To date, there have been no long-term studies in HMO's, although they are a major treatment provider in both the private and public sectors. We build on critical findings from our study to date: we found important differences in the course of problems by gender, age, psychiatric comorbidity, and type of dependence that suggest the need for greater attention to questions about use of medical and psychiatric services, and NA/AA and social networks. We also found patterns in health disparities, particularly how loss of insurance influences the course of problems overtime. Drawing from models of health services utilization and treatment careers, we study explanatory factors: individual/characteristics (demographic, substance use, psychiatric, and medical conditions, and other social functioning), treatment characteristics (number and patterns of substance abuse treatment episodes and services received), and extra-treatment characteristics (medical and psychiatric services, recovery-oriented social networks, and self-help affiliation). We use this model to address questions such as: Do multiple treatment episodes build on one another? Does regular primary care reinforce treatment benefits over time? Will the younger group show the same maturation effects over 11 years seen in community samples? How does the course of substance use affect medical conditions? What are the predictors of mortality? How valid are self reports of substance use and medical utilization? The study tests "real world" use of substance abuse, medical, and psychiatric services and has important implications for developing a continuing care approach to treatment.