IN previous experiments in this laboratory, the egg output of mixed populations of wild-type and black Tribolium ccrstaneum Herbst adults differed significantly from Hardy-Weinberg expectations at the b locus based on the gene frequencies of the parents; in cultures with a low gene frequency of black (qb) the deviations were in favor of b, and in cultures with a high qb-in favor of 4- (SOKAL and SONLEITNER 1968). Identification of wild type (++), black (bb) and their hybrid (+b) was in the adult stage (their recognition in the egg or larval stages is at present impractical). The following factors, one or more of which could be operating at any one time, may have caused the deviations from expectation: (1 ) Nonrandom mate selection and/or nonrandom mating in the population (2) Differential fitnesses of gametes and zygotes, prior to oviposition. (3) Differences in net fecundity of females in various mating combinations. (4) Differential hatchabilities of the several types of eggs. (5) Selective forces operating on the larvae and pupae. The objective of this study was to investigate the relative importance of the first four processes as possible causes for the deviations from expectation. Factor (5) has been analyzed before in great detail ( SOICAL and SONLEITNER 1968) and was excluded from this study, with precautions taken in experimental design to minimize its effect. The four factors were investigated in reverse order, starting with the hatching eggs (offspring) and working backward to their parents. One reason for this was the feeling that it would be preferable not to challenge the assumption of random mating unless no alternative explanation can be found for observed deviations from expectation. In population genetic and evolutionary studies, zygotic frequencies of off spring are generally predicted from the gene frequency of the parents on the basis of the Hardy-Weinberg law in organisms as far apart as Drosophila and man (e.g., LEVITAN 1954; TRANKELL 1955; PREVOSTI 1964). For such predictions to be valid, the population should be infinitely large, randomly mating, and free from with respect to the b locus, or linked loci.