This study investigated the time course and magnitude of the pharmacokinetic interaction between capecitabine and the cytochrome P450 (CYP) 2C9 substrate celecoxib, with implications for coadministration of fluoropyrimidines with CYP2C9 substrates such as warfarin. Patients received celecoxib 200 mg orally twice daily continuously, with capecitabine (1000 mg/m2 orally twice daily for 14 days every 21 days) starting 7 days later. Assessment of the drug‐drug interaction (DDI) potential was performed using equivalence testing, which assumes that there is no clinically relevant DDI when the calculated 90% confidence intervals (CIs) of the drug exposure ratios fall within the range of 0.80 to 1.25. Comparison of steady‐state pharmacokinetic parameters of celecoxib between day 7 (cycle 0, celecoxib only) and day 14 (cycle 1, celecoxib + capecitabine) showed geometric mean ratios of 1.24 (90%CI, 1.04‐1.49), 1.30 (1.11‐1.53) and 1.28 (1.11‐1.47) for maximum plasma concentration, minimum plasma concentration, and area under the concentration‐time curve from time zero to 8 hours, respectively. Comparison of day 7 vs day 21 (cycle 1, after 1 week washout of capecitabine) showed a further increase in the geometric mean ratio of maximum plasma concentration (1.39; 90%CI, 1.16‐1.66), minimum plasma concentration (1.53; 1.10‐2.12) and area under the concentration‐time curve from time zero to 8 hours (1.41; 1.19‐1.68). Because the 90%CIs fell outside the prespecified equivalence margin, we conclude that coadministration results in a DDI (increased celecoxib exposure) that persists for at least 7 days after capecitabine discontinuation. Close monitoring should be undertaken when administering fluoropyrimidines with CYP2C9 substrates with narrow therapeutic indexes while also weighing the benefits and risks for individual patients.