Abstract: The bioactivity of a working reference standard was determined by replicate bioassays with calibration against a primary reference standard. In this study the number of bioassay replicates needed for calibration first was calculated theoretically, and if the mean value of the experimental bioassay replicates fell within the predefined bioactivity level the bioactivity of the working reference was defined as 100%. Our results showed that when the total intermediate precision of the bioassay method was at 11.66% and the predefined bioactivity level was set at 95%-105% with a confidence level of 95%, 21 bioassay replicates should be carried out for calibration. The average value of the 22 experimental bioassay replicates was 101.96%, so the bioactivity of the working reference standard was consistent with that of the primary reference standard at 100%. The results suggest that a strategy of first calculating the number of bioassay replicates needed for calibration and then determining whether the resulting experimental mean value is within the predefined bioactivity level will be of value to the biopharmaceutical industry.