To determine whether CYP19A1 polymorphisms are associated with abnormal activity of aromatase and with musculoskeletal and bone side effects of aromatase inhibitors. DNA was isolated from tumor specimens of 4861 postmenopausal women with hormone receptor-positive breast cancer enrolled in the BIG 1–98 trial to receive tamoxifen and/or letrozole for 5 years. Tumors were genotyped for six CYP19A1 polymorphisms using PCR-based methods. Associations with breast cancer-free interval (BCFI), distant recurrence-free interval (DRFI), musculoskeletal and bone adverse events (AEs) were assessed using Cox proportional hazards models. All statistical tests were two-sided. No association between the CYP19A1 genotypes and BCFI or DRFI was observed overall. A reduced risk of a breast cancer event for tamoxifen-treated patients with rs700518 variants was observed (BCFI CC/TC vs. TT: HR 0.53, 95 % CI 0.34–0.82, interaction P = 0.08), but not observed for letrozole-treated patients. There was an increased risk of musculoskeletal AEs for patients with rs700518 variants CC/TC versus TT (HR 1.22, 95 % CI 1.03–1.45, P = 0.02), regardless of treatment. Tamoxifen-treated patients with rs4646 variants had a reduced risk of bone AEs (AA/CA vs. CC: HR 0.76, 95 % CI 0.59–0.98), whereas an increase of minor allele (C) of rs10046 was associated with an increased risk of bone AEs (HR 1.28, 95 % CI 1.07–1.52). rs936308 variants were associated with a reduced risk of bone AEs in letrozole-treated patients (GG/GC vs. CC: HR 0.73, 95 % CI 0.54–0.99), different from in tamoxifen-treated patients (GG/GC vs. CC: HR 1.32, 95 % CI 0.92–1.90, interaction P = 0.01). CYP19A1 rs700518 variants showed associations with BCFI, DRFI, in tamoxifen treated patients and musculoskeletal AEs regardless of treatment. SNPs rs4646, rs10046, and rs936308 were associated with bone AEs.