The Polyp Prevention Trial was designed to evaluate the effects of a high-fiber (18 g/1,000 kcal), high-fruit and -vegetable (3.5 servings/1,000 kcal), low-fat (20% energy) diet on recurrence of adenomatous polyps. Participants > or =35 years of age, with histologically confirmed colorectal adenoma(s) removed in the prior 6 months, were randomized to the intervention or control group. Demographic, dietary, and clinical information, including use of nonsteroidal anti-inflammatory drugs (NSAID), was collected at baseline and four annual visits. Adenoma recurrence was found in 754 of 1,905 participants and was not significantly different between groups. NSAID use was associated with a significant reduction in recurrence [odds ratio (OR), 0.77; 95% confidence interval (95% CI), 0.63-0.95]. In this analysis, NSAIDs modified the association between the intervention and recurrence at baseline (P = 0.02) and throughout the trial (P = 0.008). Among participants who did not use NSAIDs, the intervention was in the protective direction but did not achieve statistical significance (OR, 0.87; 95% CI, 0.69-1.09). The intervention was protective among males who did not use NSAIDs at baseline (OR, 0.71; 95% CI, 0.54-0.94), but not among NSAIDs users (OR, 1.09; 95% CI, 0.74-1.62). For females, corresponding OR estimates were 1.28 (95% CI, 0.86-1.90) and 2.30 (95% CI, 1.24-4.27), respectively. The protective association observed for NSAID use was stronger among control (OR, 0.63; 95% CI, 0.47-0.84) than for intervention group participants (OR, 0.97; 95% CI, 0.74-1.28). These results should be interpreted cautiously given that they may have arisen by chance in the course of examining multiple associations and Polyp Prevention Trial study participants were not randomly assigned to both dietary intervention and NSAID use. Nevertheless, our results suggest that adopting a low-fat, high-fiber diet rich in fruits and vegetables may lower the risk of colorectal adenoma recurrence among individuals who do not regularly use NSAIDs.