The possible influence of induced abortion on breast cancer risk was assessed in a cohort of 267 040 women enrolled in a randomised trial of breast self-examination in Shanghai, China. Based on answers to a baseline questionnaire, subsequent breast cancer risk was not significantly associated with ever having an induced abortion. After adjustment for potential confounders, the relative risk estimate was 1.06 (95% C.I.: 0.91, 1.25), and there was no trend in risk with number of abortions. Analysis of data from more detailed interviews of 652 cases and 694 controls from the cohort yielded similar results. There was also no overall increase in risk in women with induced abortion after first birth. Few women had undergone an abortion after 13 weeks gestation or before their first child. Although increases in risk were observed in such women, they were not statistically significant and could have been due to recall bias. Abortions as they have been performed in China are not an important cause of breast cancer.