It is argued that there is some confusion in the literature concerning the effect of assortative mating on dominance variance. It is shown that in Fisher's model of assortative mating additive deviations of one factor are correlated with additive deviations of another and that this correlation is responsible for the increase in genetic variance. The dominance deviations of one factor are independent of the dominance, as well as of the additive deviations of a second factor. It is shown that in Fisher's model of assortative mating and to the approximation used by him, the dominance variance does not change under assortative mating.