Social protection programs are needed more than ever during periods of social upheaval but are also likely to be even harder to implement successfully. Furthermore, social upheaval makes measuring the impact of such policies all the more difficult. We study the impact of a multi-faceted social protection program, often referred to as a “graduation” model program, in Yemen during a period of civil unrest. We are unable to measure outcomes for four years, thus much remains unknown about what transpired in the intermediary time. After four years we find positive impacts on asset accumulation and savings behavior, albeit substantially less than the amount the household originally received.