This investigation was a randomized controlled trial to determine the impact of health education interventions on the return of mailed fecal occult blood (FOB) tests (FOBT adherence) in a colorectal cancer screening program. The study sample included 2,201 men and women aged 50 to 74 years who were members of an Independent Practice Association (IPA)-type health maintenance organization (HMO). Subjects were randomly assigned to a "usual care" Control Group (advance letter, screening kit, reminder letter), and Treatment Groups 1 (usual care + reminder call), 2 (usual care + self-held screening booklet + reminder call), or 3 (usual care + self-held screening booklet + instruction call + reminder call). Bivariate analysis revealed significant differences in adherence (P less than .001) across study groups: Control Group (27%), Group 1 (37%), Group 2 (37%), Group 3 (48%). In addition, a significant positive association between age and adherence (P less than .001) was found. Logistic regression analysis revealed an interaction between sex and treatment. Adherence among men in all treatment groups increased significantly (P less than .0001) in relation to Control Group males. Men in Group 3 also were more likely to adhere than those in Group 2 (P less than .01) or Group 1 (P less than .01). Among women, adherence was significantly higher in Group 3 than in Group 2 (P less than .03), Group 1 (P less than .025), or the Control Group (P = .0008). The primary reason cited for nonadherence was perceived inconvenience of the FOB testing procedure.