Disaster aid is often too little, too late. Pressure on aid budgets is prompting donors to find ways to handle more crises with less funding. But the current model of discretionary, ex-post disaster aid is increasingly insufficient for these growing needs, and does little to create incentives for governments in affected countries and donors to invest in risk reduction and resilience. This framing paper sets out how the global community can do better. It proposes combining novel insurance contracts that provide fast payouts based on ‘parametric’ triggers with clear incentives to manage risks and invest in reducing losses. Pilot programmes show that the model can work. The challenge is to provide coverage for a broader range of risks, explicitly align incentives in managing risks and reducing losses, and determine the appropriate role for governments and donors. Properly implemented, an insurance paradigm for disaster aid will save many more lives for much less money.