Background: NCIC CTG MA.27 compared adjuvant steroidal [exemestane (E)] and non-steroidal [anastrozole (A)] aromatase inhibitors (AIs) and showed neither to be superior in breast cancer outcomes. AIs are known to increase the risk of osteoporosis. We examined the effects of baseline or subsequent self reported osteoporosis on disease outcomes. Methods: MA.27 enrolled 7,576 women. Event free survival (EFS) was the primary endpoint. Distant disease-free-survival (DDFS) was a secondary outcome. MA.27 permitted bisphosphonates to prevent or treat osteopenia or osteoporosis unless prohibited by enrollment to one of two cohorts in a bone substudy. Osteoporosis was considered present for this analysis if reported at baseline or prior to relapse or breast cancer death. Multivariate stratified Cox regression was used to examine the effects of trial therapy, osteoporosis, baseline patient and tumour characteristics on EFS; DDFS; bone only relapse; bone concurrent with other relapse; and non-bone recurrence. Bone marrow recurrence (N=8) was excluded. Results: Osteoporosis was reported at baseline by 654 of 7576 (8.6%) women, and prior to relapse by an additional 661 women. EFS events occurred in 693/7576 (9.15%). Osteoporosis was significantly associated with better EFS [HR 0.81 (95% CI 0.66-0.99), p=0.04] with no difference in multivariate HR of E vs A: 1.02, 95% CI 0.88-1.19, p=0.77. Women experienced 313 (4.1%) DDFS events. Osteoporosis was associated with better DDFS [HR 0.71 (95% CI 0.51-0.98), p=0.04], adjusted HR of E to A: 0.94 (95% CI of 0.75-1.17), p=0.56. Both EFS and DDFS interactions of osteoporosis with trial therapy were not significant (p>0.05). Osteoporosis was not significantly associated with the site of relapse. Conclusions: Osteoporosis had a significant prognostic association with improved EFS and DDFS in women treated with AIs. We plan to investigate the use of bisphosphonates, raloxifene treatment prior to randomization, and markers of bone resorption with outcome.