We use both molecular and experimental tools to study plant mating systems and floral evolution. The mating system can vary from obligate outcrossing to complete selfing. The mating system is a prime factor in explaining differences in levels of genetic variation among populations and species. Limited variation, as is often found in selfing populations, may strongly reduce the evolutionary potential of selfers. Current work focuses on three issues: 1) The molecular evolution of the self-incompatibility locus, 2) The evolution of mating system diversity in higher plants. 3) Floral variation, its molecular causes and selective consequences.