Despite a massive research effort, our understanding of why, in most vertebrates, males compete for mates and females care for offspring remains incomplete. Two alternative hypotheses have been proposed to explain the direction of causality between parental care and sexual selection. Traditionally, sexual selection has been explained as a consequence of relative parental investment, where the sex investing less will compete for the sex investing more. However, a more recent model suggests that parental care patterns result from sexual selection acting on one sex favoring mating competition and lower parental investment. Using species-level comparative analyses on Tanganyikan cichlid fishes we tested these alternative hypotheses employing a proxy of sexual selection based on mating system, sexual dichromatism, and dimorphism data. First, while controlling for female reproductive investment, we found that species with intense sexual selection were associated with female-only care whereas species with moderate sexual selection were associated with biparental care. Second, using contingency analyses, we found that, contrary to the traditional view, evolutionary changes in parental care type are dependent on the intensity of sexual selection. Hence, our results support the hypothesis that sexual selection determines parental care patterns in Tanganyikan cichlid fishes.