Various hypotheses about adaptive and non-adaptive mechanisms of non-parental infant care have been put forward for different taxa (Emlen et al., 1991). The Neotropical callitrichid primates are renowned for their cooperative care of the twin litters. None of the studies conducted in the wild included information on genetic relationships within groups. This, however, is indispensable to evaluate the relevance of competing hypotheses concerning direct or indirect fitness gains. We studied two groups of wild moustached tamarins with known genetic relationships over a one-year period to examine individual time-budgets and contributions to infant carrying and food-transfer. With these data we tested whether helping behaviour might be a non-adaptive trait and, if not, whether indirect benefits via kin-selection could be excluded as an evolutionary force maintaining it. Other hypotheses on direct fitness benefits were discussed as far as (anecdotal) data permitted. Changes in time-budgets suggest costs, thus clearly refuting hypotheses assuming non-adaptivity. High within-group relatedness suggests kin-selection to be one driving force of maintaining the trait. However, non-parental individuals may help despite low relatedness. Data were not sufficient to decide which possible direct benefits most likely play a role in inducing non-relatives to help. Yet, two (non-exclusive) explanations seem to be the most probable ones: The chance to inherit the main-breeding position, and a certain chance of own direct reproductive success (the latter only for male helpers) due to polyandrous mating by the female. Other adaptive mechanisms may enhance benefits but are unlikely to be major selective forces since fitness gains are presumably rather small or uncertain.