A case-control study was conducted to assess the role of diet in the etiology of colon cancer. Diet was measured by means of a comprehensive quantifiable food frequency history instrument in 246 cases and 484 controls drawn from the general population of Utah. Each subject's diet was described by major nutrient groups and total energy based on the nutritional content of foods reported. Cases reported higher daily food intake 5 years preceding diagnosis than controls [men, rate ratio (RR) = 2.5; women, RR = 3.6], as measured by total energy content of the diet. Higher risk of colon cancer with increasing energy intake was independent of stage of disease at diagnosis and obesity, as measured by body mass. Fat, protein, and carbohydrate intake all had elevated RRs but could not be assessed as risk factors independent of energy intake because of their strong correlations with total calories. Due to the higher energy intake of the cases, odds ratios for the daily intake of dietary fiber and vitamins A and C were also greater than 1. However, adjusting for caloric intake removed this effect, and dietary fiber showed a weak protective effect. Total energy intake must be evaluated before attempting to assign a causal role to any food or nutrient that may be postulated to play a role in colon cancer.