Few studies have examined the association between physical activity and disease outcomes in breast cancer survivors. Here, we report the association between lifetime total physical activity performed prior to diagnosis and disease outcomes in a population-based sample of breast cancer survivors. A cohort of 1,231 women diagnosed with breast cancer between 1995 and 1997 was followed for a minimum of 8.3 years for any cancer progressions, recurrences and new primaries; and a minimum of 10.3 years for deaths. All treatment and follow-up care received was abstracted from medical records. Data on physical activity including type (occupational, household, recreational) and dose (frequency, intensity and duration) performed during the entire lifetime until diagnosis were examined in Cox proportional hazards models as well as with cumulative incidence curves. An average of 126 MET-hr/week were reported for total physical activity (of which 13.9, 46.9 and 65.3 MET-hr/week were, respectively, for recreational, occupational and household activity). A decreased risk of breast cancer death and all deaths was observed among women in the highest versus the lowest quartiles of recreational activity (MET-hr/week/year) (HR = 0.54, 95% CI = 0.36-0.79). Both moderate (0.56, 95% CI = 0.38-0.82) and vigorous intensity recreational activity (0.74, 95% CI = 0.56-0.98) decreased the risk of breast cancer death. Moderate intensity recreational activity decreased the risk of a recurrence, progression or new primary cancer (0.66, 95% CI = 0.48-0.91). No other association with breast cancer survival was observed for other types of physical activity. Prediagnosis recreational activity conferred a benefit for survival after breast cancer. Moderate intensity recreational activity was particularly protective.