Female Enallagma hageni oviposit underwater where they are inaccessible to males. I demonstrate that males guard submerged females rather than perch sites, and are behaviorally distinct from lone males at the water. In contrast to lone males, which always attempt to copulate with females presented to them, guarding males exhibit a conditional latency to remating which corresponds closely to the time required by females to oviposit a complete clutch of eggs. By ovipositing underwater, females decrease the risk that their eggs become exposed. Risks associated with submerged oviposition favor both mate guarding, and multiple, within-clutch matings by females. Both guarding mates and lone males ’rescue’ females that float on the water surface as the result of improper resurfacing. Such behavior reduces the mortality risk to females from 0.06 to 0.02 per oviposition bout. By remating between bouts, females benefit from the additional vigilance of lone males, who rescue floating females 1.4 times as often as original mates. A second consequence of multiple mating is an increase in the selective advantage of vigilance by mates. Because receptive females become scarce by early afternoon, whereas male density remains high, a mlae has little (3%) chance of encountering a second receptive female that day. However, he incurs a 42% risk of losing fertilizations if he abandons a mate. For male E. hageni mate guarding functions in the context of both natural and sexual selection. It insures that a mate lives to lay a complete egg clutch in addition to protecting a male’s sperm investment.