Capecitabine, an orally available prodrug of 5-fluorouracil, is widely used to treat patients with colorectal cancer. Although various studies have shown circadian variations in plasma 5-fluorouracil concentrations during long-term infusion, it is still unknown whether circadian variations also exist following administration of capecitabine. The present study aimed to investigate whether the pharmacokinetics of capecitabine and its metabolites, including 5-fluorouracil, vary according to administration time in rats. Rats were orally administered capecitabine (180 mg/kg) at 07:00 (23 h after light onset, HALO), 13:00 (5 HALO), or 19:00 h (11 HALO). Plasma concentrations of capecitabine and its metabolites, such as 5′-deoxy-5-fluorocytidine (5′-DFCR), 5′-deoxy-5-fluorouridine (5′-DFUR), and 5-fluorouracil, were determined after capecitabine administration. The results showed that the t1/2 and AUC0–∞ values of 5-fluorouracil differed as a function of the dosing time of capecitabine. The maximum and minimum mean t1/2 values of 5-fluorouracil were obtained when the drug was administered at 07:00 h (23 HALO: 3.1 ± 1.2 h) and 13:00 h (5 HALO: 1.5 ± 0.6 h), respectively. The AUC0–∞ value of 5-fluorouracil at 07:00 h (23 HALO: 533.9 ± 195.7 μmol·h/L) was 1.8-fold higher than the value at 13:00 h (5 HALO: 302.5 ± 157.1 μmol·h/L). The clearance of 5-fluorouracil followed a cosine circadian curve, and the simulated population mean clearance was highest at rest times and lowest during active times in rats. The results for the plasma 5′-DFCR and 5′-DFUR levels indicated that circadian variations in the sequential metabolism of capecitabine to 5-fluorouracil would also affect plasma 5-fluorouracil levels following capecitabine administration. In conclusion, the pharmacokinetics of capecitabine and its metabolites, including 5-fluorouracil, varied according to time of dosing, suggesting that the capecitabine administration time is an important factor in achieving sufficient efficacy and reducing toxicity in patients.