Apolipoprotein E (apoE) plays a major role in the metabolism of bile acids, cholesterol and triglycerides, and has recently been proposed as being involved in the carcinogenic process. Given the potential role of bile acids in colorectal cancer etiology, it is reasonable that colorectal cancer risk might be modified by apoE genotype. We used data collected from a case-control study of colon cancer (n=1556 cases and 1948 controls) and rectal cancer (n=777 cases and 988 controls). The absence of an e3 apoE allele significantly increased the risk of colon cancer (OR=1.37 95% CI 1.00-1.87), particularly among those diagnosed when older than 64 years (OR=1.88 95% CI 1.17-3.04; P interaction between age and apoE genotype equal to 0.05). A significant three-way interaction was detected for family history of colorectal cancer, age at diagnosis and apoE genotype (P = 0.05), in those diagnosed when older, not having an e3 allele and having a significantly increased risk of colon cancer with family history of colorectal cancer (OR=3.93 95% CI 1.23-12.6). This was compared with the risk associated with family history of colorectal cancer among those diagnosed when older, with an e3 allele of 1.61 (95% CI 1.17-2.23) or those diagnosed when younger without an e3 allele (OR=2.40 95% CI 0.56-10.3). Among those diagnosed when older than 64 years, associations of BMI and prudent diet with colon cancer were stronger among individuals without an e3 allele, although the P for interaction was not significant. We did not detect any significant associations between apoE genotype and rectal cancer, survival after diagnosis with colorectal cancer, stage of disease at diagnosis or type of tumor mutation. These findings suggest those apoE genotypes that do not include the e3 allele, the same genotypes that are associated with increased risk of coronary heart disease, may influence development of colon cancer among those who are older at diagnosis.