Factors related to women's practice of breast self-examination (BSE) were studied in a sample of upper-middle-class women attending meetings of voluntary women's groups. Factors found to relate directly to frequency of BSE practice were high perceived benefits of BSE, low perceived barriers to BSE, and high self-concept. Correlations with perceived susceptibility/seriousness of breast cancer, age, perceived level of social support, and social network properties were not significant. A multiple regression analysis was done with BSE frequency as criterion variable; perceived threat (susceptibility/seriousness) and perceived benefits minus perceived barriers were entered hierarchically; age, self-concept, and total social support were entered as a group. Perceived benefits minus barriers was the only significant predictor variable, R2 = .27. These findings underline the importance of assessing detective behaviors such as BSE as potentially different from preventive behaviors.