OBJECTIVES: Capecitabine (Xeloda®), an oral chemotherapeutic agent, is effective to treat both adjuvant and metastatic colorectal cancer (CRC) patients. In the recent X-ACT trial, capecitabine compared with intravenous 5-FU/LV adjuvant therapy demonstrated superior relapse-free survival (65.5% vs. 61.9% at three years follow-up (p=0.0407) in patients with Dukes' C colon cancer and improved covariate-adjusted overall survival (p=0.0208). Based on results from X-ACT, this study assesses the cost-effectiveness of capecitabine from both US payer and societal perspectives. METHODS: Trial-based data were collected on treatment period medical resource use. Unit costs for drug administration, hospitalizations, consultations, concomitant medications, and patient time were imputed using published sources. A health-state transition model was used to estimate incremental cost impact and effectiveness in terms of the gains in relapse-free months (RFMs) (mean trial follow-up: 3.8 years) and life years (LYs). Future outcomes were discounted at 3% annually. RESULTS: Mean duration of treatment was approximately 24 weeks in both arms. Administration of capecitabine required fewer clinic visits per patient (7.4 vs. 28.0 with 5-FU/LV). Mean acquisition costs of capecitabine were $8700 higher than with 5-FU/LV, but these costs were partially offset by 5-FU/LV's administration cost of $5700. Total hospital days for treatment-related adverse events (AEs), medication costs for treating AEs, and patient time costs were higher for 5-FU/LV. The cost of physician consultations for treating AEs did not differ. Compared to 5-FU/LV, capecitabine increases RFMs by 1.3 months at four years yielding a cost-effectiveness ratio of $2500 per RFM gained (payer perspective); from a societal, lifetime perspective, capecitabine increases LYs by 11.3 months with a cost per LY gained of $500. This finding was not sensitive to plausible variations in key parameters. CONCLUSIONS: Based on our model, capecitabine, an oral chemotherapeutic agent, is cost-effective in the US compared with intravenous 5-FU/LV in treating patients with adjuvant CRC.