Infanticide in non-human primates occurs when an individual kills its own or another individual's dependent young. Five hypotheses have been proposed to explain infanticide in non-human primates: exploitation, resource competition, parental manipulation, sexual selection, and social pathology.

Infanticide in non-human primates occurs as a result of exploitation when the individuals performing the infanticide directly benefit from consumption or use of their victim. The individual can become a resource: food (cannibalism); a protective buffer against aggression, or a prop to obtain maternal experience.

The form of exploitation in non-human primates most attributable to adult females is when non-lactating females take an infant from its mother (allomothering) and forcibly retain it until starvation. This behavior is known as the "aunting to death" phenomenon; these non-lactating female primates gain mothering-like experience, yet lack the resources to feed the infant. This behaviour has been seen in captive bonobos, but not wild ones. It is not clear if it is a natural bonobo trait or the result of living in captivity. Male orangutans have not been directly observed practicing infanticide as a reproductive strategy, but recorded case of a male abducting an infant almost resulting in said infant dying from dehydration was observed. Additionally, a possible case of infanticide has been inferred, in which a mother orangutan had lost an infant and received a serious injury on her foot shortly after a new male had been introduced nearby. Although not directly observed, it is inferred this male attacked the female and killed her infant.

Resource competition results when there are too few resources in a particular area to support the existing population. In primates, resource competition is a prime motivator for infanticide. Infanticide motivated by resource competition can occur both outside of and within familial groups. Dominant, high ranking, female chimpanzees have been shown to more often aggress towards a lower ranking female and her infant due to resource competition. Primates from outside of familial groups might infiltrate areas and kill infants from other groups to eliminate competition for resources. When resources are limited, infants are easier to eliminate from the competition pool than other group members because they are the most defenseless and thus become targets of infanticide. Primate infanticide motivated by resource competition can also involve cannibalizing the infant as a source of nutrition.

Resource competition is also a primary motivator in inter-species infanticide, or the killing of infants from one species by another species. Through eliminating infants of another species in the same environment, the probability that the aggressor and their own infants will obtain more resources increases. This behavior has been an observed consequence of multiple primate inter-species conflicts. In these cases, instances of direct aggression toward inter-specific infants in addition to infanticide have also been observed. In these instances of direct aggression, the aggressor was the previous target of intra-species aggression directed towards them. Therefore, the direct aggression and infanticide carried out by these aggressors could be attributed to re-directed aggression.

Maternal infanticide, the killing of dependent young by the mother, is rare in non-human primates and has been reported only a handful of times. Maternal infanticide has been reported once in brown mantled tamarins, Saguinus fuscicollis, once in black fronted titis, Callicebus nigrifrons, and four times in mustached tamarins, Saguinus mystax. It is proposed that maternal infanticide occurs when the mother assesses the probability for infant survival based on previous infant deaths. If it is unlikely that the infant will survive, infanticide may occur. This may allow the mother to invest more in her current offspring or future offspring, leading to a greater net reproductive fitness in the mother.

In the instances of maternal infanticide in tamarins, there were multiple breeding females. The parental manipulation hypothesis proposes that maternal infanticide occurs more frequently when the group has a poor capacity to raise offspring, multiple breeding females, birth intervals shorter than three months, and low infant survival probability.

Maternal infanticide differs from other varieties of infanticide in that the resource competition and sexual selection hypotheses (see other sections) must be rejected. Resource competition and sexual selection are ruled out because it is the mother that is performing the infanticide, not another female.