It is widely agreed that the flowers of hermaphrodite plants evolve in response to selection acting simultaneously through male and female sexual functions, but we know very little about the pattern of gender-specific selection. We review three current hypotheses for gender-specific selection by viewing them within a single phenotypic selection framework. We compile data from phenotypic selection and manipulative studies and evaluate the fit between empirical data and the hypotheses. In this preliminary analysis, we find that neither the male-function hypothesis nor the gender-balance hypothesis is well supported. However, the context-dependence hypothesis is supported by the documented diversity of gender-specific selection and by evidence that selection through female fertility is significantly correlated with pollen limitation of seed production. Future studies contributing to our understanding of selection through male and female function in plants need to quantify and manipulate the ecological context for reproduction, as well as describe male and female fitness responses to fine-scale trait manipulation.