Forty-nine litter series sired by Tt[degree], Tt1, Tt3, Tt9, Tt12, and Ttp1 males of the house mouse were examined for heterogeneity of the transmission ratios of the t-allele. It was found that heterogeneity often exists, not only between males with the same t-allele, but even between litters from the same male; the latter was demonstrated by means of a new statistical technique. To measure the degree of heterogeneity, the mean square contingency was used as an index. It was found that intra-male heterogeneity was generally lower with "late" mating than with "normal" mating. It appears that the behavior in the female tract of spermatozoa carrying a t-allele differs from that of spermatozoa that do not have such an allele, and, further, that this behavior is sensitive to slight variations in the physiological conditions in the tract. The statistical method used for examining the significance of the intra-male heterogeneity is described in an Appendix. It is a Monte Carlo method necessitating the use of an electronic computer. The method is useful in situations where cell frequencies are too low for the application of ax2 test.