Male gametophytic selection can play a special role in the evolution of higher plant populations. The main assumption — gametophytic-sporophytic gene expression of a large portion of a plant's genes — has been proven by a number of studies. Population analyses have revealed a large amount of variability for male gametophytic fitness. However, the data available do not prove that at least a portion of this variability is due to postmeiotic gene expression. This paper reports the analysis of a synthetic population of maize based on a gametophytic selection experiment, carried out according to a recurrent scheme. After two cycles of selection, the response was evaluated for gametophytic and sporophytic traits. A parameter representing pollen viability and time to germination, although showing a large amount of genetic variability, was not affected by gametophytic selection, indicating that this variability is largely sporophytically controlled. Pollen tube growth rate was significantly affected by gametophytic selection: 21.6% of the genetical variability was released by selection. Correlated response for sporophytic traits was observed for mean kernel weight: 15.67% of the variability was released. The results are a direct demonstration that pollen competitive ability due to pollen tube growth rate and kernel development are controlled, to a considerable extent, by genes expressed in both tissues. They also indicate that gametophytic selection in higher plants can produce a higher evolution rate than sporophytic selection; it can thus serve to regulate the amount of genetic variability in the populations by removing a large amount of the genetic load produced by recombination.