DESCRIPTION (provided by applicant): This is an R21 application for innovative/developmental research by an experienced substance dependence investigator. The study of psychological characteristics of patients that are predictive of response to substance abuse treatment has largely focused upon negative characteristics such as depression, anxiety, antisocial personality, or hostility, i.e., psychological illness. Although a growing literature has shown that various dimensions of psychological wellness of individuals and patients are associated with positive health outcomes in general, there has been only limited investigation of whether aspects of psychological wellness are also predictive of response to substance abuse treatment. The proposed research is designed to determine whether psychological strengths such as optimism, personal meaning, perceived wellness, spirituality/religiosity, positive mood, or social support are predictive of a more positive response to treatment in 300 substance abuse patients and whether the benefits of these psychological strengths are additive or interactive. Outcomes predicted will be treatment completion and during-treatment cocaine urine toxicologies; and days of cocaine use and heavy alcohol use during study months 4-6. The research further inquires whether the contribution of psychological wellness to the prediction of treatment outcomes is independent of that predicted by psychological illness and includes statistical evaluations for two mediation hypotheses which may help to explain the effects of optimism and spirituality/religiosity. The research will also implement a within- and between-instrument latent structure analysis of the various measures of psychological strengths in order to determine their underlying dimensionality. A number of secondary analyses will be undertaken. These include determination of the predictive validity of salient sociodemographic variables such as gender, evaluation of the predictive validity of psychological wellness for other relevant outcomes (e.g., days of heroin use, psychological and legal status at follow-up; risk for AIDS behaviors), evaluation of the relationships between psychological wellness and illness, and exploratory recursive partitioning analyses which may provide information on complex combinations of predictor variables.