Since the start of the transition, the Russian Federation has experienced a number of regulatory and administrative reforms with respect to the provision child benefits. Besides changes in financing from local to federal budget, the introduction of means testing of previously universal child benefits was the most radical reform in child benefit provision. The current programme has been in place since 2000. Using the cross-section and panel components of the Russia Longitudinal Monitoring Survey (RLMS) from 2000 to 2004, this paper is the first to evaluate whether the targeting efficiency of child benefits has increased during this period and whether benefit receipt adequately assists households in the prevention of (chronic) poverty. We show that incidence and coverage rates under the target population have increased strongly but that leakage remains considerable. Our analysis also shows that the impact of child benefits on family welfare is rather small and benefit receipt helps little to prevent (chronic) poverty. Simulations of four alternative benefit scenarios show that universal child benefits perform slightly better in terms of poverty reduction. Although all tested scenarios perform better than the current scheme, doubling the size of the universal benefit and allocate it universally, would have the largest impact and reduce poverty with 12% at a cost of an estimated 0.025% of GDP.