In this study, we assessed the pharmacokinetics (PK), bioequivalence, and safety of 150 mg capecitabine compared to the branded reference formulation in colorectal or breast cancer patients receiving a high-fat diet. This was a multicenter, open, random, balanced, three-period, three-sequence and semi-repetitive cross study with 48 subjects. In each study period, the eligible subject received the test or reference formulation, followed by a 1-day washout period. Serial blood samples for pharmacokinetic assessment were collected at predose up to 8 h postdose. The plasma concentrations of capecitabine were analyzed by LC/MS–MS. Pharmacokinetic parameters (non-compartmental model) were assessed with WinNonlin software. The pharmacokinetic parameters assessed were the area under the plasma concentration-time curve from time 0 to the time of last measurable concentration (AUC0–t), the AUC from time zero to infinity (AUC0–∞), the peak plasma concentration of the drug (Cmax), the time needed to reach maximum concentration (Tmax), the elimination half-life (t1/2), and the terminal elimination rate (λz). All were analyzed using an analysis of variance (ANOVA) model after logarithmic transformation of the data. To establish the bioequivalence (BE) for capecitabine, reference-scaled average bioequivalence (RSABE) acceptance criteria and average bioequivalence (ABE) acceptance criteria were used. Safety and tolerability were assessed during the entire study period. Reference scaled maximum plasma concentration (Cmax) was higher than 0.294, permitting use of RSABE. The within-subject SDs of the reference intervention (SWR) for AUC0–t and AUC0–∞ were < 0.294, meeting ABE criteria. The point estimate for the geometric least squares mean (GLSM) ratio for the point estimate of Cmax was 0.962, within the range of 0.80–1.25. The 90% upper confidence boundary for the test/reference of GLSM ratios was 97.84–105.40% for AUC0–t and 97.33–103.51% for AUC0–∞, all of which were within the prespecified limits. The 90% confidence intervals for AUC0–t and AUC0–∞ and 95% upper confidence limit for Cmax indicated bioequivalence. No serious adverse events were found among the subjects. According to the criteria for bioequivalence, the test formulation was bioequivalent to the reference formulation in terms of the rate and extent of absorption under fed conditions by measurement of total capecitabine and was safe and well tolerated. NCT04420871.