This study involved field and laboratory investigations of three species of the genus Dineutes in an attempt to find out if distribution patterns (in Michigan) suggest that interspecific competition is relevant, and to determine possible mechanisms of competition. Broad areas of overlap exist in the continental ranges of the three species. Seasonal aspects of their life cycles are similar with two exceptions. Dineutes nigrior and D. assimilis begin breeding slightly earlier than D. horni, and D. assimilis has a considerably shorter breeding season than the other two species. Egg production and larval growth are very similar in all three species. Whirligig beetles seem to be relatively free from predation. The field study yielded a number of results. (1) Relative abundances of three species are correlated with north-south position. In southern Michigan, horni is dominant, but nigrior is dominant to the north. Dineutes assimilis is nearly always rare and disappears in the northernmost lakes. (2) On a local scale, in southern Michigan, horni is dominant in lakes whereas nigrior and assimilis are the usual dominants in ponds. (3) Such habitat exclusion is most marked during the breeding season. (4) Two or three of the species are known to have completed reproduction simultaneously in some localities, constituting cases of coexistence. A number of points were established by the laboratory experiments. (1) Dineutes larvae are cannibalistic and the intensity of individual larval aggression is inversely related to the amount of food received. (2) Single species populations of larval Dineutes adjust numerically to the existing food supply through cannibalism. (3) Less food per individual is required to suppress cannibalistic behavior in nigrior than in horni, and assimilis is still less responsive to food level. (4) In interspecific competition between nigrior and horni, horni is the better competitor. Whenever the adjustment of the total number of larvae of both species is just low enough to reduce cannibalistic behavior in nigrior it is still insufficient to reduce such behavior in horni. The result is a disproportionate number of nigrior deaths. (5) Lower temperature and a high initial relative abundance of nigrior reduces the advantage held by horni. (6) In competition between assimilis and nigrior, assimilis usually wins but the outcome can be reversed by lower temperature and higher initial density. It appears than in nature regulation of population size by larval cannibalism would be an adaptation for dealing with a variable food supply. Patterns of distribution, the general natural history of Dineutes, and the likelihood that larval populations are resource-limited make it reasonable that competition does occur in nature, although rigorous proof is lacking. In places where coexistence occurs it may result from small-scale (small pond) environmental intermediacy such that no species is uniquely favored in competition. Heterogeneity of the environment on a larger scale may supply neighboring places where each of the species is uniquely favored. The result of these two possibilities would presumably be the observed overall pattern of habitat exclusion with occasional cases of coexistence. It is proposed that interspecific competition acts as a selective force during coexistence and that this function of competition is distinct from the idea of competitive exclusion.