Family history data of colorectal cancer, heart disease, and stroke were obtained on near relatives (parents, siblings, and children) in 702 colorectal cancer cases and 710 age-/sex-matched community controls as part of a large, comprehensive, population-based epidemiological and clinicopathological study of colorectal cancer conducted in Melbourne (the Melbourne Colorectal Cancer Study). There was a statistically significant higher family history rate of colorectal cancer in cases than in controls (relative risk = 2.13; 95% confidence interval = 1.53-2.96; p less than 0.001). This family history effect was more pronounced for colon cancer than for rectal cancer and there was an earlier age of detection of colorectal cancer in those with a family history of this cancer when compared with those without such a history. Dietary risk factors for colorectal cancer, which were previously described in the Melbourne study, were separate and independent from the family history effects. It is concluded that a family history of colorectal cancer is an important indication to screen individuals for this cancer, and also that while heredity has a definite role in the etiology of colorectal cancer, this hereditary effect is either likely to be small, or else likely to be important in only a proportion (perhaps 20%) of cases.