This study evaluates the effect of treatment components through a secondary analysis of data from the National Treatment Improvement Evaluation Study (NTIES). The study examines the relationship between treatment components, client-level factors, and treatment outcomes, and how these relationships vary by treatment modality. It seeks to understand the response by NTIES clients to components of drug and alcohol treatment defined in ways that have external applicability and that are immediately available to care providers. Several dose components were shown to have significant effects on the odds of positive outcome. Several interactions between client-level factors and dose components were observed. Compared to the background variables alone, the inclusion of treatment components significantly improved the logistic regression models for non-methadone outpatient and short-term residential modalities both for outcomes and for methadone for treatment substance abstinence. Improvement for the long-term residential modality was marginal and not significant for the correctional modality. Implications for research, policy, and practice are discussed. Appendix A is "Description of the National Treatment Improvement Evaluation Study and Center for Substance Abuse Treatment Demonstrations (1990-1992)"; Appendix B is "Distributions of SDU Means for Treatment Components within Modalities"; and Appendix C is "Technical Notes." (Contains 10 figures, 8 tables, and 21 references.) (MKA)