In this thesis, I demonstrate how sociological data can be used to test predictions derived from evolutionary psychology, sensu lato. In the first section of the thesis, I test hypotheses from kin selection theory. Firstly, I test the hypothesis that relatedness influences adult sibling ties. Using a large scale dataset on family relationships from the Netherlands (NKPS), I show that both maternal and paternal half-siblings differ significantly from full siblings in sibling tie strength. In a follow up study, I also show that basic knowledge about a sibling, namely whether the sibling is still alive or not, varies according to sibling type. If siblings are not fully related they are less likely to have any basic knowledge about each other. Subsequently, I test the role of birth order for family relationships and specifically sibling ties. Kin selection theory suggests, that else being equal, firstborns will invest more in laterborn siblings, rather than vice versa. Using the same set (NKPS), I show support for this prediction. Unlike studies from undergraduate populations, however, I fail to find any evidence for a 'neglected middleborn' effect. Middleborns did not significantly differ from other birth orders in their sibling ties or their relationships with family members. The findings on birth order studies are discussed with reference to the current literature on kin competition, birth order and family dynamics. In the next part of the first section, I analyze how.childlessness influences family relationships. Kin selection theory suggests that childless individuals will be more inclined to invest in their siblings with children, than vice versa. I find support for this prediction in a sample of Belgian women and a sample from an historical American population. The findings of these studies are discussed with reference to the helpers-at-the-nest hypothesis. In the final two studies of this first section I test whether or not lineage affects of grandparental investment in kin. Using botha Dutch sample and a British sample, I show, that all else being equal, maternal grandparents, especially, maternal grandmothers, invest more in their grandchildren than paternal grandparents. The results are discussed with reference to current research on paternity uncertainty. In the second section of the thesis, I test predictions derived from the evolutionary psychology of mate choice, and life history theory. Firstly, I demonstrate the effect of a biological market on human mate choice using an historical sample. If there is an oversupply of men on the marriage market then women could demand higher status from their partner. This prediction from biological market theory was supported. In the following study, I test the hypothesis that female orgasm frequency is related to male quality using a sample of Chinese women. I find that male wealth, and to a lesser extent male height, predict female orgasm frequency. This is consistent with an adaptationist view of the female orgasm. In the final part of the second section, I present two studies testing hypotheses derived from life history. In the first study, I demonstrate that in a stressed environment, rural Guatemala, 'growing tall' is an accurate predictor of maternal fitness. In the second study, I evaluate the generalised Trivers-Willard Hypothesis which predicts that taller and bigger women will have relatively more sons than smaller and thinner women. I fail to find any support for this hypothesis in British and rural Guatemalan data. I conclude by briefly discussing why the use of a wide variety of secondary data resources covering modern societies is important for evolutionary psychology.