ABSTRACT Data are presented on the reliability of the previously described bioassay of gonadotrophin neutralizing potency of antisera raised against HCG preparations. The sensitivity of the method may be defined as the smallest detectable neutralizing activity expressed in anti-units (AU). This depends partly on the bioassay system used, and partly on the intended percentage of neutralization. In the HCG bioassay used in the present study, and around 50 % neutralization, the smallest detectable neutralizing activity is about 1.0 AU. The sensitivity decreases with increasing levels of neutralization: it is about 9 AU around 90% neutralization. The precision as given by the mean index of precision (λ) was 0.12 in 52 assays. The 95 % fiducial limits expressed as a percentage of the mean potency estimate ranged from 65.8 % to 125.7% in 52 assays conducted at 50 % and from 89.1% to 105.3% in 25 assays conducted at 90% intended neutralization. Reproducibility: eighteen antisera raised in 18 rabbits against 9 HCG preparations were assayed at 50% intended neutralization repeatedly (at least twice each antiserum) in 52 assays. Only one antiserum showed significant heterogeneity among repeated estimates by a χ2 test. The achieved neutralization ranged from 36 % to 79 % with a mean of 50 %. The neutralizing potency estimates of 19 antisera raised in 19 rabbits against 10 HCG preparations in 45 assays at 50% intended neutralization and in 25 assays at 90 % intended neutralization were also compared. These potency estimates were in close agreement with each other at both levels of neutralization (except for one antiserum) as evidenced by the results of χ 2 test and the value of the correlation coefficient obtained (r = 0.99). It is concluded that "the principle of simple additivity" is a satisfactory approximation to represent the antigen - antibody reaction involved in the neutralization of the biological activity of HCG preparations.