Colorectal adenomas and possibly some hyperplastic polyps are precursors of colorectal cancer. Tobacco use is associated in epidemiologic studies with these polyps, although links between smoking and colorectal cancer are less consistent. To characterize the role of tobacco in early colorectal carcinogenesis, we compared tobacco use among 4,383 subjects with histologically verified benign (hyperplastic or adenomatous) polyps of the distal colon (descending colon, sigmoid, and rectum) with tobacco use among 33,667 subjects who were endoscopy negative for distal colon tumors, in the screening arm of the Prostate, Lung, Colorectal, and Ovarian Trial, a randomized trial of flexible sigmoidoscopy. Risks, estimated by the odds ratio (OR), associated with current cigarette use were OR = 4.4 [95% confidence interval (95% CI), 3.7-5.2] for hyperplastic polyps only, OR = 1.8 (95% CI, 1.5-2.1) for adenomas only, and OR = 6.2 (95% CI, 4.7-8.3) for subjects with both hyperplastic and adenomatous polyps concurrently. Effects were weaker among ex smokers; the smoking-associated ORs remained consistently higher for hyperplastic polyps. This pattern was also seen in relation to cigarettes smoked per day, smoking duration, and pack-years. Tobacco-associated risks for multiple polyps were also stronger when hyperplastic disease was involved. In conclusion, tobacco use, particularly recent use, increases risk for both adenomatous and hyperplastic polyps, but the risks are substantially greater for hyperplastic lesions.