A central concern with humanitarian assistance is that unconditional cash transfers might discourage work. Yet refugees differ from the chronically poor in most cash-transfer studies: many retain skills that transfers might help reactivate. We study the world's largest humanitarian cash transfer scheme, the Emergency Social Safety Net (ESSN) in Turkey, which long hosted more refugees than any other country. Using a survey representative of the refugee population, we exploit the program's dependency-ratio eligibility rule in a regression discontinuity design. Far from discouraging work, cash transfers increase the probability that any man in the household works by 13.2 and the share of working men by 19.3 percentage points. We find no evidence of an effect on women's already-low labor supply. A design-based local-randomization analysis with exact finite-sample inference corroborates these effects (9.8 and 17.4 percentage points). Transfers also improve access to information, reduce reliance on harmful food-coping strategies, and raise Turkish-language course enrollment and education spending. These patterns suggest transfers relax the liquidity and information constraints refugees face rather than inducing a shift toward leisure.