We use a set of experiments to study the effects of forced military service for a rebel group on social capital. We examine the case of Northern Uganda, where recruits did not self-select nor were systematically screened by rebels. We find that individual cooperativeness robustly increases with length of soldiering, especially among those who soldiered during early age. Parents of ex-soldiers are aware of the behavioral difference: they trust ex-soldiers more and expect them to be more trustworthy. These results suggest that the impact of child soldiering on social capital, in contrast to human capital, is not necessarily detrimental.