Trastuzumab has conferred significant clinical benefits in HER-2-positive breast carcinomas. HER-2 status is determined by immunohistochemistry (IHC) and/or fluorescence in situ hybridisation (FISH), but appropriate assessment of HER2 status remains subject to considerable debate. Data on the health economic impact of HER-2 test strategies are limited. A life-long Markov state transition model was used to assess costs and effectiveness of HER-2 assay strategies (based on IHC, FISH, both combined or FISH confirmation of IHC2+) for a hypothetical cohort of early breast cancer patients from the perspective of the Swiss health system. We compared clinically relevant strategies of predictive testing and subsequent trastuzumab treatment of HER-2-positive patients only. FISH testing was the most cost-effective strategy with an incremental cost-effectiveness ratio of <euro>12,245 per additional quality-adjusted life-year (QALY) gained, compared to no trastuzumab treatment. The next best strategy was parallel IHC and FISH, with costs of <euro>400,154/QALY gained compared to FISH alone. FISH as primary HER-2 testing modality remained the preferred option in deterministic and probabilistic sensitivity analysis. Predictive testing to identify adjuvant breast cancer patients who benefit from trastuzumab treatment is a clinical and economic necessity. Our model identifies FISH as the most cost-effective approach.