Three aspects of mating behavior—time to mating, repetitiveness, and mate selection—were investigated in six genetic strains of the flour beetleTribolium castaneum and their hybrids. The data were compared with predictions from theoretical models. Time to mating was found to be heritable and characteristic of a strain. In homotypic strain combinations the duration of mating was longer than in heterotypic combinations. In repeated matings the male mating rate declined with time. In a multiple-choice experiment there was a tendency for positive assortative mating; this did not occur in female-choice or male-choice tests. Keeping beetles of two strains together prior to the experiment removed this tendency. Males almost exclusively preferred virgin to fertilized females when given a choice. The data fitted Taylor's [(1975).Behav. Genet. 5:381–393] model better than Kence and Bryant's [(1978).Am. Nat. 112:1047–1062] model. The absence of assortative mating when males (and females) of two strains were held together can be explained by the pheromone-saturation model of Averhoff and Richardson [(1974).Behav. Genet. 4:207–225] but not by Bryant's [(1979).Behav. Genet. 9:249–256] alternative.