This article describes the details underlying the targeting of a Nicaraguan anti-poverty program, emphasizing the rationale for how it was designed and implemented. It offers, by way of example, a guide for targeting in an anti-poverty program, and highlights some of the potential tradeoffs. It then goes on to present a quantitative assessment of how well the program was able to target poor households. A combination of ad hoc and statistical procedures led to targeting that was effective, with undercoverage rates of 10 percent or below and leakage rates of 15 percent or below. This was in spite of the fact that the targeting methodologies used were imprecise at both the household and geographic levels.