ABSTRAcr.-Data obtained from published studies and from mice trapped for use in our laboratory were analyzed to determine the relationship between frequency of t-haplotypes in wild house mice (Mus musculus) and other variables such as sex, age at capture, and kind of t-haplotype found in the population. In addition, the frequency of t-alleles in populations from which mice had been sampled repeatedly was examined to determine whether, within populations, frequency of t-haplotypes tends to change over time. Populations in which semilethal haplotypes predominated had higher frequencies of heterozygous animals than populations in which the predominant haplotype was a lethaL Furthermore, frequencies of t-haplotypes were higher in males than in females. However, no difference in frequency of t-alleles was found between mice trapped as juveniles and individuals of the same sex trapped as adults. Furthermore, only one of 14 populations from which mice were trapped repeatedly showed evidence of a change in frequency of heterozygotes over time. The relevance of these data for potential processes controlling the frequency of t-haplotypes in wild populations is discussed.Considerable debate (Lacy, 1978) has existed over factors controlling the frequency of alleles at the T-locus in wild house mice (Mus musculus). About 25% of wild house mice are heterozygous (+ It) for a variable recessive allele and the remainder usually are homozygous (+ I +) for the wild-type allele (Bennett, 1978). All recessive t-alleles (or t-haplotypes) are deleterious when homozygous. Some alleles (designated "semilethal") produce some viable homozygotes but cause sterility in homozygous males. Others ("lethal alleles") cause death in all homozygous embryos. These lethal haplotypes can be classified further by complementation testing into 16 groups (Klein et aI., 1984) of which two (t wi and t wS ) are common in North American wild house mice (Bennett, 1978).The principal factor maintaining these deleterious alleles in wild populations is transmissionratio distortion in heterozygous males that may transmit their t-haplotype to as many as 95-100% of their progeny. The debate over factors controlling the frequency of t-haplotypes in natural populations has arisen because theoretical models incorporating selection against homozygotes counteracted by transmission-ratio distortion yield predicted equilibrium frequencies of t-carrying individuals substantially higher than those actually observed (Bruck, 1957; I?,u.nn and Levene, 1961; Lewontin, 1968). Hypotheses proposed to account for the lower thun.expected frequency of t-haplotypes have included inbreeding (Petras, 1967), genetic drift (Le·wontin· andDunn, 1960), interdemic selection (Lewontin, 1962), heterozygote disadvantage (Drickamer and Lenington, 1987;Egid and Lenington, 1985;Johnston and Brown, 1969;Lenington, 1983;Lenington and Egid, 1985; Levine et aI., 1980;Meyers, 1973) and lowered transmission ratio of t-haplotypes in natural matings as compared with laboratory test crosses (Lenington, 19...