AbstractWe distinguish the effects of the Philippines’ conditional cash transfers (CCT) between the intended (poor) and unintended (nonpoor) beneficiaries. We use a proxy means test similar to what was officially used and apply it on a nationally representative sample of households to identify some poor households who are excluded and nonpoor households included in the program on account of inherent leakages in program implementation. To evaluate the CCT’s impact, we matched the poor beneficiaries with poor households excluded from the program and nonpoor beneficiaries with nonpoor households outside CCT. For the poor households, we find a reduction in wages of working children who presumably now spend more time in school in compliance with the program's conditionalities, indicating less child labor. For nonpoor households, we find some indication that program participation led to moral hazard as evidenced by lower income from salaries and wages. Since the reduction in income is greater than the increase in transfers received by nonpoor households, their savings declined.