This meta-analysis examined 118 studies of the psychosocial outcomes of children exposed to interparental violence. Correlational studies showed a significant association between exposure and child problems (d = -0.29). Group comparison studies showed that witnesses had significantly worse outcomes relative to nonwitnesses (d = -0.40) and children from verbally aggressive homes (d = -0.28). but witnesses' outcomes were not significantly different from those of physically abused children (d = 0.15) or physically abused witnesses (d = 0.13). Several methodological variables moderated these results. Similar effects were found across a range of outcomes, with slight evidence for greater risk among preschoolers. Recommendations for future research are made, taking into account practical and theoretical issues in this area.