We examined the relationship between adult stature and cancer incidence using data from the first U.S. National Health and Nutrition Examination Survey and its follow-up study. Among 12,554 participants 25-74 years old, 460 cancers occurred in men and 399 in women after an average follow-up period of approximately 10 years. The age-adjusted relative risk of cancer for the second (Q2) through fourth (Q4) quartiles of stature compared to the first quartile among men were significantly increased: 1.5, 1.4, and 1.4. After adjustment for race, cigarette smoking, income, and body mass index, the all-sites cancer relative risk increased slightly to 1.6, 1.5, and 1.6. For most cancer sites in men, and particularly colorectal cancer (relative risk = 2.1 for Q4), the lowest incidence was observed among those in the shortest quartile of stature. A weaker, positive association was evident among women, restricted primarily to cancer of the breast and colorectum (relative risk in Q4 = 2.1 and 1.6 for the two cancers, respectively). These findings indicate that short stature is associated with reduced risk of cancer, particularly in men, and suggest a role for nutrition early in life in human carcinogenesis.