Scholars of the resource curse argue that reliance on primary commodities destabilizes governments: price fluctuations generate windfalls or periods of austerity that provoke or intensify conflict. 350 quantitative studies test this claim, but prominent results point in different directions, making it difficult to discern which results reliably hold across contexts. We conduct a meta-analysis of 46 natural experiments that use difference-in-difference designs to estimate the causal effect of international commodity price changes on armed conflict. We show commodity price changes, on average, do not change conflict risks. However, this overall effect comprises cross-cutting effects by commodity type. In line with theory, we find price increases in labor-intensive agricultural commodities reduce conflict, while increases in the price of oil, a capital-intensive commodity, provoke conflict. We also find that prices changes for lootable artisanal minerals provoke conflict. Our meta-analysis consolidates existing evidence, but also highlights gaps for future research to fill.