Systematic observations have revealed that cancer patients experience a wide array of emotional reactions. Because of limited helping resources, as well as the possibility of prevention, it would be useful to identify those cancer patients at risk for the development of significant psychiatric difficulties and psychosocial maladjustment. Many investigators have proposed, but not substantiated factors which seem to be associated with such problems following cancer diagnosis. This report reviews clinically noted or theoretically-derived factors which have been tested empirically for relationships with various aspects of psychosocial adjustment. Certain specific cancer sites have been noted to be associated with psychosocial problems. A specific biological basis for psychiatric problems associated with certain diseases has been proposed for multiple myeloma, lung tumors and pancreatic cancer. A number of chemotherapy agents are now recognized as accounting for presumed psychiatric symptoms. While studies relating age, sex, marital or socioeconomic status with psychosocial problems have found no consistent associations, social support has been demonstrated as a significant factor in adjustment to cancer. Studies of the relation of adaptation and the patient's psychological situation such as degree of pessimism or anxiety, personality characteristics, prior experience with the death of a close friend or relative or religious beliefs have been inconclusive. Many studies have used measures of questionable external validity and others employed unvalidated subjective ratings. It is only recently that there are instruments considered reliable and valid for research in this area and even these instruments have not been extensively validated. Finally, this paper reviews several major design problems which have impaired identification of predictive variables and provides suggestions for future research.