many benthic marine invertebrates develop by means of free-livlng, dispersive larval stages. The presumed advantages of such larvae include the avoidance of competition for resources with adults, temporary reduct~on of benthic mortality while in the plankton, decreased likelihood of inbreeding in the next generation, and increased ability to withstand local extinction However, the direct~on of evolutionary change appears generally b~a s e d toward the loss of larvae in many clades, implying that larvae are somehow disadvantageous. Poss~ble disadvantages include dispersal away from favorable habitat, mismatches between larval and luvenile physiological tolerances, greater susceptib~lity to env~ronmental stresses, greater susceptibihty to predation. and vanous costs that may be associated with n~etanlorphosing in response to specific chemical cues and postponing n~etamorphosis in the absence of those cues. Understanding the forces responsible for the present distribution of larval and non-larval (aplanktonlc) development among benthic marine invertebrates, and the potential influence of human activities on the direct~on of future evolutionary change in 1-eproductlve patterns, will require a better understanding of the following issues. the role of macro-evolutionary forces in selecting for or against dispersive larvae, the relative tolerances of encapsulated embryos and free-living larvae to salinity, pollutant, and other environmental stresses; the degree to which egg masses, e g g capsules, and brood chambers protect developing embryos from environmental stresses; the relative magmtude of predation by planktonic and benthic predators on both larvae and early juveniles; the way In which larval and juvenile size affect vulnerability to predators; the extent to w h~c h encapsulation and brooding protect against predators; the amount of genetic change associated with loss of larvae from invertebrate life cycles and the time required to accomplish that change; and the degree to which continued input of larvae from other populations deters selection against dispersive larvae The prominence of larval development in modern life cycles may reflect difficulties In loslng larvae from llfe cycles more than selection for their retention.