We conducted a randomized controlled trial among African-American patients attending a primary-care provider visit to compare efficacy of a computer-delivered tailored intervention to increase colorectal cancer (CRC) screening (n = 273) with non-tailored print material—an American Cancer Society brochure on CRC screening (n = 283). Health Belief Model constructs were used to develop tailored messages and examined as outcomes. Analysis of covariance models were used to compare changes between CRC knowledge and health belief scores at baseline and 1 week post-intervention. At 1 week, patients who received the computer-delivered tailored intervention had greater changes in CRC knowledge scores (P < 0.001), perceived CRC risk scores (P = 0.005), FOBT barriers scores (P = 0.034) and colonoscopy benefit scores (P < 0.001). Findings show that computer-delivered tailored interventions are an effective adjunct to the clinical encounter that can improve knowledge and health beliefs about CRC screening, necessary precursors to behavior change.