Epidemiologic studies investigating the effects of nonsteroidal anti-inflammatory drugs (NSAIDs) on breast cancer have yielded conflicting results. We examined the association between use of aspirin and nonaspirin NSAIDs and breast cancer risk among 28 695 women in the Danish Diet, Cancer and Health cohort. Information on NSAID and paracetamol use was obtained from a self-administered questionnaire completed at baseline (1993-1997) and updated through 2003 using a nationwide prescription database. Detailed information on breast cancer incidence and tumour characteristics was obtained from nationwide health registers. Cox proportional hazards regression was used to compute incidence rate ratios (RRs) and 95% confidence intervals (CIs). We identified 847 breast cancer cases over an average follow-up period of 7.5 years. Any NSAID use at baseline was associated with an increased incidence of breast cancer compared with nonuse (RR, 1.27; 95% CI, 1.10-1.45). A similar result was observed for any NSAID use in a combined analysis of baseline and prescription data (1.34; 95% CI, 1.15-1.56). Aspirin-only users experienced a slightly higher breast cancer incidence (RR, 1.38; 95% CI, 1.12-1.69) than exclusive users of nonaspirin NSAIDs (RR, 1.25; 95% CI, 1.04-1.49). Introduction of a lag time of 1 year provided similar results. We found no clear differences in risk estimates with frequency, recency or duration of NSAID use, or by hormone receptor status of the breast tumours. Paracetamol use was unrelated to breast cancer incidence. The increased breast cancer incidence among NSAID users may reflect a noncausal association, but our study provides no evidence of a chemopreventive effect of NSAIDs against breast cancer over the durations studied.