OBJECTIVES: The 76 months follow-up of the BIG 1-98 study reported a hazard ratio [HR] for overall survival (OS) of 0.81 and a HR for disease free survival (DFS) of 0.84 in the letrozole (LET) group versus tamoxifene (TAM). The 100 months follow-up of the ATAC study reported almost  identical results observed at 68 months follow-up for the HR of OS (0,97) in the anastrozole (ANA) group when compared with TAM. The HR for DFS was 0.85. This analysis compares the cost-effectiveness of LET and ANA versus tamoxifen from the Canadian healthcare perspective using the most recent survival data from the BIG 1-98 and ATAC studies. METHODS: A Markov model was used to estimate the cost per quality adjusted life year (QALY) of LET or ANA versus TAM. Annual survival probabilities were extracted from updated BIG 1-98 and ATAC trials, and literature. Cost values representing resource use were informed by a Canadian costing analysis. Utility weights were derived from literature. A time horizon of 20 years and a 5% discount rate were used. RESULTS: LET and ANA are predicted to increase QALYs by 0.26 and 0.05 per patient compared to TAM, respectively. Lifetime costs increase by $2710 and $2297 for LET and ANA respectively.  The incremental cost-effectiveness ratio (ICER) of LET versus TAM is $10,420. The ICER for ANA versus TAM is $41,569. The model is more sensitive to the variation of the hazard ratios for overall survival. The probability that LET and ANA are cost-effective given a threshold value of 50,000 per QALY is 99% for LET and 58% for ANA. CONCLUSIONS: Our model shows that LET has a substantially lower expected ICER of $10,420 compared to ANA, with mean ICER of $41,569 over a 20 year time horizon from the Canadian healthcare perspective.