Females of facultatively eusocial insects are all capable of laying eggs and therefore all have the potential to practice different reproductive strategies. Direct reproduction can be achieved by being dominant in a group or by nesting alone. Indirect reproduction is achieved by helping to raise the offspring of relatives. The choice of strategy depends on how reproductive conflicts are resolved and this may be determined by genetic and ecological factors. In this thesis I examine the factors affecting reproductive conflicts in hover wasps (Stenogastrinae, Hymenoptera). Using highly polymorphic microsatellite markers, developed by screening partial genomic libraries, I estimate relatedness among colony members in three species of Stenogastrinae (Parischnogaster alternata, Liostenogaster flavolineata and L. vechtii). Relatedness in P. alternata was significantly lower than in L. flavolineata. I attribute the low relatedness in P. alternata to the movement of females between nests, their tendency to found nests in groups and that they often nest with females from other groups. Liostenogaster flavolineata was previously perceived as only being incipiently eusocial because of low relatedness among female nestmates (r = 0.22 0.1 - Strassmann et al 1994). I re-examine colony fine genetic structure in this species. By genotyping brood and adults, I show that there is normally only one reproductive laying female eggs, but that subordinates occasionally lay male eggs. I conclude that reproductive strategies in L. flavolineata are typical of a eusocial insect. I discuss the parameters from reproductive skew models that may explain the high skew. The little variation in skew amongst colonies of L. flavolineata meant that the models could be evaluated only at the interspecific level. A comparative analysis across 21 hymenopterans revealed a positive relationship between relatedness and skew. This result is consistent with a prediction of the Incentives model for reproductive skew (Reeve and Ratnieks 1993).