Although the killing of dependent infants by adult males is a widespread phenomenon among primates, its causes and consequences still remain hotly debated. According to the sexual selection hypothesis, infanticidal males will gain a reproductive advantage provided that only unrelated infants are killed and that the males increase their chances of siring the next infants. Alternatively, the social pathology hypothesis interprets infanticide as a result of crowded living conditions and, thus, as not providing any advantage. Based on DNA analyses of wild Hanuman langurs (Presbytis entellus) we present the first evidence that male attackers were not related to their infant victims. Furthermore, in all cases the presumed killers were the likely fathers of the subsequent infants. Our data, therefore, strongly support the sexual selection hypothesis interpreting infanticide as an evolved, adaptive male reproductive tactic.