This paper systematically reviews the work on the antipoverty effect of SNAP using administrative and survey data. We found the antipoverty effects are even larger than those found in Census Bureau estimates if one adjusts for underreporting. Using re-weighting methods to benchmark the CPS to administrative data, the antipoverty effects are then almost again as large as without them. With underreporting adjustments, and depending on the poverty measure being considered, SNAP reduces poverty by 14 to 16 percent. We conclude that SNAP is our nation’s most effective antipoverty program for the non-elderly when adjusted for underreporting, and it is especially good at reducing extreme poverty—by over 50 percent—and also especially effective for poor families with children.