A reanalysis of treatment response and relapse was performed using survival analysis in a 12-week clinical trial of cognitive behavioral group psychotherapy for the treatment of bulimia nervosa. One hundred forty-three (143) bulimic women with high incidence of binge eating, self-induced vomiting, and/or laxative abuse were randomly assigned to one of four possible treatment conditions that consisted of a combination of two factors: (1) emphasis on abstinence (high and low), and (2) treatment intensity (high and low). "Initial" and "maintained" response to treatment based on "total" and "near" abstinence criteria were determined using self-reported binge eating, vomiting, and laxative use data. Results suggest that an emphasis on abstinence appears important in achieving initial abstinence, whereas intensity of treatment may be important in maintaining abstinence.