In a case control study which formed one arm of a large, population-based investigation of colorectal cancer incidence, aetiology and survival. 'The Melbourne Colorectal Cancer Study', among others, 22 psychosocially orientated questions were asked by personal interview of 637 histologically confirmed new cases of colorectal cancer and 714 age/sex frequency matched community controls, from Melbourne (population 2.81 million). Self-reported childhood or adult life 'unhappiness' was statistically significantly more common among the cancer cases, while 'unhappiness with retirement' was similarly distributed among cases and controls. Questions which were formulated to test a particular personality profile as a cancer risk, and which included the elements of denial and repression of anger and of other negative emotions, a commitment to prevailing social norms resulting in the external appearance of a 'nice' or 'good' person, a suppression of reactions which may offend others and the avoidance of conflict, showed a statistically significant discrimination between cases and controls. The risk of colorectal cancer with respect to this model was independent of the previously found risk factors of diet, beer intake, and family history of colorectal cancer, and was also independent of other potential confounding factors of socioeconomic level, marital status, religion and country of birth. Although the results must be interpreted with caution, the data are consistent with the hypothesis that this personality type may play a role in the clinical expression of colorectal cancer and merits further study.