OBJECTIVE The purpose of this study was to assess the utility of HBM as a theoretical guide for predicting breast cancer screening and therefore for guiding intervention studies. BACKGROUND Breast cancer is the leading cause of death for middle age women (35-50) and the second leading cause of cancer deaths in all women in the United States (US). Early detection of breast cancer through screening is the only option available to women. However, less than half of all women in the US participate in screening. The health belief model (HBM), which specifies interactions of values and beliefs about health and their influence on choices, has been widely used to explain screening behaviour. METHODS An integrative research review analysed 16 published descriptive studies employing HBM. Literature was located through a search of research based studies listed in Cumulative Index of Nursing and Allied Health (CINAHL), Medline, and cancer literature databases and studies cited in other references between 1990 and 1999. RESULTS Application of HBM was inconsistent. No study tested nonlinear relationships between variables as specified in the model. At best, the model explained 47% of the variance in screening behaviour when socioeconomic status was included. Otherwise predictive power was low, ranging from 15 to 27%. CONCLUSIONS While the model provides some description of the values, beliefs and behaviours of middle-aged women primarily, HBM does not appear to have the power to consistently predict behaviours. Further research is needed to provide more thorough depiction of the social, nonhealth care meaning of breast cancer.