In this study two important factors that are thought to govern interspecific variation in pollen-ovule ratios were examined. First, the effect of habitat disturbance on variation in pollen-ovule ratio was determined. The second factor studied was the pollination type, used as a surrogate for the efficiency of pollination. Because seed mass is known to be strongly correlated with the pollen-ovule ratio it was also included in the analyses to examine if a possible effect of habitat disturbance or pollination type is still valid after accounting for the effect of seed mass. Furthermore, phylogenetically comparative methods were used to investigate whether the correlations between traits were maintained through evolutionary history or are only present in recent species data, i.e. in analyses that do not consider phylogenetic relationships between species. In conflict with the reproductive assurance hypothesis, habitat disturbance did not have a significant effect on interspecific pollen-ovule ratio variation. In contrast, pollination type accounted for a significant proportion of the variation in pollen-ovule ratios, even after taking into account the strong effect of seed mass. General results do not differ between the cross-species and phylogenetic comparative approaches. The results both accord with the predictions of the sex allocation theory and the proposition that the chance of a pollen grain reaching a stigma governs the pollen-ovule ratio.