OBJECTIVES: To evaluate the sensitivity of the cost per QALY of trastuzumab for the treatment of EBC using one-way and probabilistic sensitivity analysis (PSA). METHODS: A 5-state Markov model with annual transition cycles was constructed to estimate the long-term health outcomes of EBC patients based on results from the HERA clinical trial. Population-based utilities were used for the health states in the model. NHS resource use and costs were estimated from a consensus panel of experts and published unit costs respectively. Costs and benefits were discounted at 3.5% per annum. Using the base case Markov model, key assumptions and model parameters were varied through plausible ranges identified in the literature to evaluate the stability of the base case cost per QALY. Key assumptions modified in the sensitivity analysis included: 1) proportion of patients receiving trastuzumab in the metastatic setting in both the Adjuvant trastuzumab and No adjuvant trastuzumab arms when they develop metastatic disease; 2) baseline patient age; 3) the duration of the treatment effect of trastuzumab; 4) the baseline risk of disease progression. Five thousand iterations were applied in the PSA with beta pert and beta distributions applied to the ranges of each parameter. RESULTS The estimated base case cost per QALY for adjuvant trastuzumab was low at £2387. Trastuzumab remained cost effective in all evaluated scenarios with the cost per QALY always falling below the cost effectiveness threshold of £30,000. The parameters with the largest impact upon the cost per QALY were the re-treatment rate in the metastatic setting and the duration of the treatment effect. 100% of all iterations were below £30,000 per QALY in the PSA. CONCLUSION The cost per QALY of trastuzumab for the treatment of EBC has been demonstrated to be robust and remain below commonly accepted thresholds despite wide variations in model assumptions.