Sixty-six patients with Duke's B2 or C colon or rectal cancer were randomized for treatment with aspirin (ASA), 600 mg, p.o., twice daily x 2 years or placebo (P). Compliance was checked in both groups by random measurement of blood salicylate levels. Fifty-seven patients are currently evaluable. No difference in disease-free (p = .66) or overall survival (p = .90) is present between ASA and P groups. The time at which ASA therapy is started (within 2 or within 4 weeks) following surgery does not affect these results. Aspirin at conventional dosage is ineffective in preventing the appearance of metastases in patients with colo-rectal cancer.