USDA’s Supplemental Nutrition Assistance Program (SNAP) provides food and nutrition benefits to low-income households based on a formula that adjusts the benefit amount a household receives based on monthly need. This study assesses the extent to which SNAP reaches the poorest households, also known as benefit targeting, by estimating benefit receipt by annual household income relative to poverty. To conduct this analysis, ERS linked 2008-12 SNAP administrative records from New York State to data from the U.S. Census Bureau’s American Community Survey (ACS), a primary source for national, State, and local information on the use of SNAP and other public assistance programs. The linked data provide better information on SNAP receipt than that which would be estimated by the ACS alone and permit a more complete characterization of SNAP targeting. Study findings show that the assessment of SNAP targeting varies by the three measures of benefit receipt examined. Estimates of SNAP targeting toward low-income households improve when using either of two measures of intensity of SNAP participation relative to measures of ever-in-the-year participation. Replacing survey-based data on SNAP benefit receipt with administrative records of SNAP benefit receipt and adjusting the survey households to more closely reflect administrative SNAP units also improves estimates of targeting to low-income participants. Finally, estimated program targeting is improved by removing sample households that do not report income and for which income is imputed in the survey from the analysis.