BACKGROUND: Postmenopausal women with osteoporosis had a 66% relative risk reduction for invasive breast cancer over 8 years of raloxifene therapy in the randomized, placebo-controlled 4-year MORE (Multiple Outcomes of Raloxifene Evaluation) trial and the CORE (Continuing Outcomes Relevant to Evista) trial, a 4-year follow-up to MORE. PATIENTS AND METHODS: The first post hoc analysis examined the effects of raloxifene on the cumulative incidence of invasive breast cancer on a yearly basis. Another analysis compared the incidence of invasive breast cancer in 3967 patients who continued raloxifene for 8 years (RLX-C, n = 2280), discontinued raloxifene after 4 years in MORE (RLX-D, n = 401), or took placebo (n = 1286) for a mean 2.9 years' treatment duration (57,338 patient-years). RESULTS: The unadjusted breast cancer incidence rate was 5.39 per 1000 patient-years in the placebo group compared with 2.26 in the RLX-C group (hazard ratio [HR], 0.41 [95% CI 0.21-0.81]) and 3.59 in the RLX-D group (HR, 0.69 [95% CI 0.23-2.01]). Because the choice of taking the study drug was not randomized in CORE, propensity scores were used to adjust for potential imbalances in baseline characteristics before CORE. Results after adjustment by the propensity score method were similar to the unadjusted results. CONCLUSION: This analysis suggests a persistent effect for breast cancer risk reduction in patients who discontinued raloxifene, although this conclusion is limited by the small sample size.