Serological diagnostics of hepatitis B in clinical laboratories is mainly based on detection of HBs antigen (HBsAg) in human serum or plasma using commercial ELISA kits. In manufacturing and laboratory practice, sensitivity of ELISA kit is measured against either international or national reference standard. This approach is necessary, but limited as it does not take into account factors that influence on HBsAg detection kit potency at low HBsAg concentration and when HBsAg subtypes/variants are present in species. Several panels containing human HBsAg positive and HBsAg negative sera were prepared and then used for testing of commercial kits HBsAg detection efficacy either under the conditions of reference or clinical laboratories: (1) broadened panels containing 138 and 157 samples were used for lot-to-lot comparison of kits detecting HBsAg; (2) low titer panel containing 21 samples was used for the evaluation the detection limit of kits; (3) mini low titer HBsAg panels containing 17 samples were used for interlaboratory control in Moscow; (4) panel containing 38 diluted human sera (35 sera with "wild" type HBsAg and 3 abnormal sera with presumably mutant HBsAg) was used for the performance evaluation of kits. Variants of HBsAg were found in 11 sera among 132 HBsAg positive sera titrated in ELISA in the presence of monoclonal antibodies against a-determinant of HBsAg. For 4 of them the measured level of HBsAg varied 10-100 times with the change of the kit or type of monoclonal antibodies. The data obtained suggest that each control panel could be a useful tool for estimation of both the kit detection potency and laboratory assay efficacy.