OBJECTIVES: This study evaluated the cost-effectiveness of first-line oral S-1 therapy, compared with standard injection therapy (taxanes), for metastatic breast cancer patients. METHODS: We performed a cost-effectiveness analysis using data from a randomized phase III trial (SELECT BC) in which 618 Japanese patients were randomly allocated to an S-1 or taxane arm. The trial demonstrated the non-inferiority of S-1 to taxane in terms of overall survival (OS) (hazard ratio, 1.05; 95% CI, 0.86–1.27, P = 0.015). Cost-effectiveness was included as a secondary endpoint of the SELECT BC trial. Therefore some of the randomized patients participated in an EQ-5D survey (N=391) and a health economic survey (N=146). The EQ-5D responses, claims, and prescription data were collected for as long as possible until death. The expected QALY obtained from both treatments was calculated using patient-level data on survival and EQ-5D, and the expected cost was calculated using patient-level survival and claim/prescription data. The analysis was performed from a public healthcare payer’s perspective. A discount was not considered because of the short time horizon (4 years). RESULTS: The estimated EQ-5D least square means and 95% CI up to 60 months were 0.748 (95% CI, 0.722–0.775) and 0.741 (95% CI, 0.713–0.769) in the S-1 and taxane arm, respectively. The expected QALY was 2.11 for the S-1 arm and 2.04 for the taxane arm, while the expected costs were JPY 5,820,000 (USD 53,000, USD 1= JPY 110) and JPY 6,130,000 (USD 56,000), respectively. S-1 was dominant with the probability of more than 55%. According to cost-effectiveness acceptability curves, probability that S-1 was cost-effective exceeded 85% when willingness-to-pay was JPY 5million (USD 45,000). CONCLUSIONS: Most data used in this analysis was derived from an empirical randomized controlled trial, enabling a robust analysis. Our results showed that oral S-1 therapy is cost-effective.