Temporal trend and migrant studies have indicated that the etiology of colorectal cancer is predominantly environmental and, hence, modifiable. Animal fat intake has been frequently, but inconsistently, associated with the risk of this disease. We conducted a population-based case-control study in Hawaii (United States) among ethnic groups at different risks of the disease to evaluate the role of dietary lipids and foods of animal origin on the risk of colorectal cancer. We interviewed 698 male and 494 female Japanese, Caucasian (White), Filipino, Hawaiian, and Chinese patients diagnosed during 1987-91 with pathologically confirmed adenocarcinoma of the colon or rectum, and 1,192 population controls matched to cases on age, gender and ethnicity. Odds ratios (OR), adjusted for caloric intake and other dietary and non-dietary risk factors, were estimated using conditional logistic regression. Intakes of total fat, saturated fat (S) and polyunsaturated fat (P) were not related to the risk of colorectal cancer. However, an inverse association was found for the P/S ratio, with ORs of 0.6 in both genders (95 percent confidence interval [CI] = 0.4-1.0 for males; CI = 0.3-0.9 for females) for the highest compared with the lowest quartile (P < 0.05 for trend). Intakes of red meat and processed meat were associated with the risk of cancer in the right colon and rectum, respectively, in men only. Fat-trimmed red meat and fish intakes were not related to risk. Chicken eaten without skin was associated inversely with risk in both genders. The strongest association was found for eggs, with an OR of 2.7 (CI = 1.7-4.0) and 2.3 (CI = 1.4-3.7) for the highest compared with the lowest quartile of intake in men and women, respectively (P < 0.001 for trend). This association was dose-dependent, not explained by known confounders or other dietary variables, and was very consistent between genders, among ethnic groups, and across all segments of the large bowel. These data suggest that the ratio of polyunsaturated to saturated fat may be a better indicator of colorectal cancer risk than the absolute amount of specific fats in the diet. They also suggest that eggs and, possibly, untrimmed red meat and processed meat increase, and chicken eaten without skin decreases, colorectal cancer risk.