The 50th anniversary of the War on Poverty in 2014 brings renewed policy attention to safety net programs aimed at alleviating poverty and its consequences. Implementation of the Affordable Care Act and its expanded options for public insurance coverage also sharpens the focus on the role of local, state and federal governments in ensuring access to means-tested programs. However, researchers interested in studying these programs and the populations they serve face a number of challenges. A pervasive issue is the availability of timely, high quality data on targeted populations and their means-tested program participation. While administrative data tracks participation it does not, by definition, track non-participants (either those eligible but not participating or those nearly eligible) nor does it typically include much socioeconomic information on participants and their families/households. Survey data - including major representative household surveys like the American Community Survey (ACS), Current Population Survey (CPS) and Survey on Income and Program Participation (SIPP) - fills the gap by collecting detailed data on employment and earnings, family structure, and other socio-demographic characteristics. All of these surveys do collect information on safety net program participation and, in some cases, benefit amount. At the same time, when compared with available aggregate administrative totals it is clear that many people do not accurately report this information. The problem of under-reporting has persisted, and for some programs, increased over time and across surveys (e.g., Wheaton 2007; Meyer, Mok and Sullivan 2009). The level and dimensions of under-reporting are better understood for long-standing surveys such as the CPS. The more recently developed ACS lacks the same breadth of research on underreporting, yet the ACS is being used extensively due to its large sample sizes and ability to generate estimates at sub-state levels and for finely specified demographic groups. The best methods for examining the issue of underreporting – and ultimately the impact and effectiveness of safety net programs – is to combine household survey data with administrative records on participation. Researchers across the country have made strides in gaining access to administrative data and, increasingly, in combining it with survey data. This panel brings together researchers from a varied set of institutions engaged in this type of work to highlight and explore advances in both methodology and understanding of ACS survey underreporting across several of the largest programs.