This thesis concentrates on evolutionary costs and benefits of insect parental care. I use phylogenetic methods to test large-scale hypotheses, and field studies to test proximate hypotheses. Initially I look at the evolution of variation in the sex performing care, reconstructing transitions across insect evolutionary history. Consistent with theory, early insects had no care, and their descendants evolved either male care, or female care followed sometimes by biparental care. Secondly, I investigate parental care trade-offs. I find that in insects, care is associated positively with offspring survival but negatively with fecundity, suggesting a general trade-off between current and future reproduction. In the second part of the thesis, I use the assassin bug genus Rhinocoris to investigate proximate costs and benefits influencing male care, the rarest form of care. High density is predicted to favour male care; I investigate why male-caring Rhinocoris live at high density on the plant Stylosanthes. Plant preference is rare in predators and I show that the plant protects eggs from predators as well as harbouring favourable prey, factors not usually linked to parental care. Lastly I investigate an unstudied sexual conflict in male-carers. If females prefer caring males, males should be selected to display their eggs conspicuously. Conspicuousness may carry costs to eggs, so females should prefer inconspicuous locations. In the field this conflict exists for one Rhinocoris species but is absent in a sister species, showing that parental care can have complex effects. My results show that while broad patterns of costs and benefits largely follow theory, finer patterns depend on subtle ecological factors.