Food-for-work (FFW) programs are commonly used both for short-term relief and long-term development purposes. This paper assesses the potential of FFW programs to reduce poverty and promote sustainable land use in the longer run. There is a danger that such programs distort labor allocation or crowd out private investments and therefore have unintended negative effects. We explore this issue using survey evidence from northern Ethiopia that we use to motivate a simple theoretical model, a more detailed version of which we then implement through an applied bio-economic model calibrated to northern Ethiopia. The analysis explores how FFW project outcomes may depend on FFW project design, market conditions, and technology characteristics. We show that FFW programs may either crowd out or crowd in private investments and highlight factors that condition whether FFW promotes or undercuts sustainable land use.