Child physical abuse is an issue of global concern. Conservative estimates set global prevalence of this type of maltreatment at 25%, its consequences and cost to society escalating with increasing frequency and severity of episodes. Syntheses of the evidence on parenting programs for reducing rates of physical abuse recidivism have, to date, not been able to establish effectiveness. Paucity of data and inconsistent inclusion criteria in past reviews made meta-analysis often impossible or uninformative. The current systematic review updates prior reviews and overcomes some of the methodological issues they encountered by pooling trial-level data from a well-defined scope of trials of parenting interventions aimed at preventing the re-abuse of children by parents with substantiated or suspected physical abuse history. Randomized controlled trials and rigorous non-randomized designs were sought via nine online databases, two trial registries, several clearinghouses and contact with experts. A total of fourteen studies of variable quality were included in this review, four of which had outcomes that enabled meta-analysis. Overall, this review presents evidence supporting the effectiveness of parenting behavioral programs based on social learning theory for reducing hard markers of child physical abuse recidivism. Meta-analysis found that the absolute risk reduction in risk of recidivism was 11 percentage points less for maltreating parents who undergo parenting programs (RD = −0.11, 95% CI [−0.22, −0.004], p = 0.043, I
2 = 28.9%). However, the pooled effect size was not statistically significant when calculated as a risk ratio (0.76, 95% CI [0.54, 1.07], I
2 = 38.4%). Policy makers and practitioners should be made aware that this intervention method is backed by promising evidence featuring modest yet significant reductions in hard markers of child physical abuse, even though the methodological robustness of these findings should be further explored in future research.