This prospective cohort study aimed to show that coffee prevents the recurrence of colorectal tumors (adenomas, precursors of colorectal cancer, and early-stage colorectal cancers) as well as colorectal cancer. The present study included 307 patients who participated in a clinical study that required endoscopy to remove a colorectal tumor. The amount of coffee consumed by the patients at study inclusion and the frequency of colorectal tumors, as detected by colonoscopy over the subsequent 4 years, were assessed. Coffee consumption was determined using a diet survey that included 3-consecutive-day food records. The risk of colorectal tumor recurrence was significantly lower (odds ratio=0.21; 95% confidence interval, 0.06–0.74) in patients who consumed more than three cups of coffee per day compared with those who consumed no coffee. No correlation was observed between the examined factors, including green tea and black tea intake and the amount of caffeine consumed. In subanalysis divided by the tumor location within the colorectum, the odds ratio of colorectal tumor recurrence in the proximal colon showed a tendency toward reduction as coffee consumption increased; however, increased coffee consumption significantly increased colorectal tumor recurrence in the distal colon. We showed that high coffee consumption reduced the overall occurrence of colorectal tumors, affected by the reduction in the proximal colon.