This study examines the FAO cash plus agriculture program in Somalia. This multi-faceted intervention provides agricultural inputs, training and cash transfers to vulnerable agro-pastoralist households living in districts and villages that experienced severe weather shocks. We exploit variations in the implementation of this program to assess the effect of receiving inputs only and inputs plus cash on a range of protective and productive outcomes. Specifically, we make use of household survey data collected in 2019 and apply a quasi-experimental Inverse Probability Weighted Regression Analysis (IPWRA) matching approach to estimate the impact of the two different interventions on food security, assets, adoption of inputs and adoption of agricultural practices. We find positive and significant impacts on a number of productive outcomes and some difference between the two treatments: while inputs seem to increase asset wealth, cash plus reduces food insecurity and higher levels of income diversification, suggesting that the cash component facilitates investments in livelihoods diversification. Moreover, we find evidence of heterogeneous impacts under conditions of weather shocks, and between socio-economic segments of the population.