Abstract: Objective：To explore whether the reversephase high performance liquid chromatography assay coupled with ultra violet (HPLC-UV) detection recommended by Japanese Society of Clinical Chemistry (JSCC) may be used as the candidate reference method, and verify the isotope dilution mass spectrometry (ID-MS), evaluate the routine methods for the determination of uric acid in human serum. Methods：The reverse-phase HPLC-UV was reproduced and its methodology characteristics was evaluated. The chromatograms of uric acid in the human serum sample treated by uricase were observed to evaluate the specificity. The linear range was observed by means of determining the concentration of a series of standard solutions of uric acid and drawing the standard curve. Randox quality control material was used to evaluate the precision of the assay. The standard solution of uric acid was used to evaluate the recovery rate. The accuracy of the reverse-phase HPLC-UV was evaluated by using the certified reference materials (SRM 909b) and the national primary reference materials GBW 09176, GBW 09175 and GBW 09174. The results determined by reverse-phase HPLC-UV were compared with those of RELA (ring trials for reference laboratories) of 2008 IFCC (International Federation of Clinical Chemistry and Laboratory Medicine Federation). The modified Bland-Altman graphical method was used to appraise the bias between the reverse-phase HPLC-UV method and 5 routine assay systems. Results：After the serum samples were treated by the uricase, the chromatographic peak of uric acid disappeared. The linear range of reverse-phase HPLC-UV assay was from 2.08 to 1 785 μmol/L. The within-run, between-run, inter-day and total coefficients of variation (CV) of reverse-phase HPLC-UV were less than 0.3%, 0.4%, 2.0% and 2.7% respectively. The average relative recovery rate was from 96% to 100.6%. The proportional bias of target values of the certified reference material SRM 909bI and Ⅱwas -2.5% and -2.3%, and the proportional bias of GBW 09176, GBW 091574 and GBW 09176 was 0.29%, -0.74% and 0.06% respectively. The proportional bias of the survey samples of RELA 2008 IFCC ring trial were 0.35% (level 1) and -0.69 % (level 2) compared with the mean values of the participating laboratories. The correlation coefficients between reverse-phase HPLC-UV and 5 routine assay systems, i.e., Hitachi, Beckman Coulter, Roche, Dade and Vitros were 0.998 9, 0.996 5, 0.999 2, 0.999 2 and 0.998 7 respectively. The deviations between of reverse-phase HPLC-UV and 5 routine assay systems were all less than biological variation (5%). Conclusion：The reverse-phase HPLC-UV developed in this study displayed as a simple, specific, rapid, precise and accurate method, and its results showed close correlation and consistency with routine traceable systems, so it can be recommended as a candidate reference method for determination of uric acid in human serum.