By determining 22,385 progeny genotypes in a test garden of Cucumis sativus containing plants homozygous for different alleles, we have shown that details of the gene flow pattern change throughout the season within a single plant population. Allele frequencies were significantly different between progeny sets from early and mid-season pollinations, rising from .014 to .019. Also, gene flow was more restricted in distance during the mid-season period. Number and density of flowers per plant and composition of the pollinator guild were very different during these two periods, a higher flower density and higher proportion of honeybees in the guild being present during the mid-season. By tagging individual flowers thoughout the season and following their fate, we saw that the percentage of flowers that produced mature fruit during the early and mid-season periods was 29.0% and 34.3%, respectively. However, seed set from flowers produced late in the season was minimal, although pollinator activity had been high and hundreds of flowers were present. Inhibition of late developing fruit by early set fruit causes this phenomenon. Consequently, late season flower density, pollen deposition patterns, or pollinator foraging behavior would be poor predictors of gene flow dynamics as the late pollen does not significantly contribute to total gene frequencies of the progeny. Estimates of potential gene flow in plant populations based on pollinator movements and activity alone cannot be accurate unless these observations are made throughout the growing season. The corresponding temporal patterns of fruit set and inhibition must be understood for the species under study, and incorporated as factors determining microevolutionary change.