ObjectiveTo study the value of hepatitis B virus surface antigen (HBsAg) in the evaluation of antiviral efficacy and its influencing factors under a complex population background resulting from various nationalities in Xinjiang. MethodsWe retrospectively analyzed patients with chronic hepatitis B (CHB) admitted and administrated with nucleot(s)ide analogues (NAs) for the first time in Traditional Chinese Medical Hospital of Xinjiang Uygur Autonomous Region from January 2012 to August 2013. The biological, virological, and serological responses were analyzed as well as the possible factors related to HBsAg levels and its reduction levels. ResultsThere were 63 CHB patients enrolled. After 48 weeks' treatment, all patients achieved biological response, and 59 of them achieved complete virological response in spite of 4 patients with partial response. In all the 30 hepatitis B virus e antigen (HBeAg) positive patients, 5 achieved HBeAg seroconversion. After correlation and regression analysis, it turned out that the history (P=0.033) and HBeAg levels at week 48 (P<0.001) were independent impact factors for HBsAg level at week 48. And the reduction degree of HBsAg at week 48 was influenced by HBsAg at week 48. In 21 patients counting to week 72 maintaining biological response, 18 achieved complete virological response. Unfortunately, all 8 HBeAg positive patients encountered no HBeAg loss or seroconversion. After correlation and regression analysis, it turned out that HBsAg level at week 72 was influenced by HBsAg at week 48 (r=0.700, P<0.001). And the decline degree of HBsAg at week 72 was related to baseline HBsAg level. ConclusionSatisfactory efficacy can be achieved via NAs treatment in CHB patients. But when HBsAg is used separately as an indicator for therapeutic efficacy, we should be aware that intrahepatic covalently closed circular DNA (cccDNA) is not only the impact factor of HBsAg variation, the history, the variations of HBeAg and HBsAg itself during the treatment should also be considered.