Microsatellite markers are widely used in population genetic analyses. However, the common occurrence of null alleles and their effect on population parameters such as observed heterozygosity, genetic differentiation coefficient, etc., needs to be addressed. Two microsatellite data sets based on original and redesigned primers were used to assess the effect of null alleles on population genetic analyses of Maire yew (Taxus chinensis var. mairei). When amplified with the original primer set, both loci exhibited the presence of null alleles. Four populations significantly deviated from Hardy-Weinberg equilibrium due to heterozygote deficiencies. Once the null alleles were eliminated using new primers, almost all populations were in Hardy-Weinberg equilibrium. Average observed heterozygosity values (He) increased from He = 0.197 to He = 0.497 while the fixation index (FIS) decreased from FIS = 0.566 to FIS = −0.060.Genetic differentiation (FST) between populations was more pronounced using the new set of primers (FST = 0.221) compared to the original one (FST = 0.150). When the original data were corrected for the presence of null alleles using the ENA method, global FST was still biased compared to the actual results estimated by the new primer set. Our empirical study on Maire yew populations suggests that null alleles can alter the results of population genetic analyses significantly. This suggests that re-designing new primers or increasing the number of markers may eliminate the effect of the null alleles.