"HALDANE's rule" states that, if species hybrids of one sex only are inviable or sterile, the afflicted sex is much more likely to be heterogametic (XY) than homogametic (XX). We show that most or all of the phenomena associated with HALDANE's rule can be explained by the simple hypothesis that alleles decreasing hybrid fitness are partially recessive. Under this hypothesis, the XY sex suffers more than the XX because X-linked alleles causing postzygotic isolation tend to have greater cumulative effects when hemizygous than when heterozygous, even though the XX sex carries twice as many such alleles. The dominance hypothesis can also account for the "large X effect," the disproportionate effect of the X chromosome on hybrid inviability/sterility. In addition, the dominance theory is consistent with: the long temporal lag between the evolution of heterogametic and homogametic postzygotic isolation, the frequency of exceptions to HALDANE's rule, puzzling Drosophila experiments in which "unbalanced" hybrid females, who carry two X chromosomes from the same species, remain fertile whereas F1 hybrid males are sterile, and the absence of cases of HALDANE's rule for hybrid inviability in mammals. We discuss several novel predictions that could lead to rejection of the dominance theory.