In recent years there has been increased focus on evaluating the efficacy of psychosocial interventions for cancer patients. Among the several limitations inherent to these programs of research, few studies have targeted patients with well-diagnosed clinical depression and little is known about factors that best predict treatment outcome and attrition in this population. Using a heterogeneous sample of cancer patients, this study was designed to explore the relations of treatment outcome and attrition with selected demographic (age, education, marital status, gender), psychosocial (pretreatment depression and anxiety, number of co-existent disorders, social support) and medical variables (health and functional status, cancer type, and length of cancer diagnosis). Results indicated that positive treatment outcome was associated with fewer coexistent anxiety disorders at pre-treatment and involvement in a marital relationship. Attrition was more likely to occur among cancer patients presenting with higher pre-treatment depression severity and those who were less educated. Implications of these findings and study limitations are discussed. (Contains 2 tables and 1 note.)