Traditionally, building‐level disaster risk reduction measures are aimed at a single natural hazard. However, in many countries the society faces the threat of multiple hazards. Building‐level disaster risk reduction measures that aim to decrease earthquake vulnerability can have opposing or conflicting effects on flood vulnerability, and vice versa. In a case study of Afghanistan, we calculate the risk of floods and earthquakes, in terms of average annual losses (AAL), in the current situation. Next, we develop two disaster risk reduction (DRR) scenarios, where building‐level measures to reduce flood and earthquake risk are implemented. We use this to identify districts for which DRR measures of one hazard increase the risk of another hazard. We then also calculate the optimal situation between the two scenarios by, for each district, selecting the DRR scenario for which the AAL as a ratio of the total exposure is lowest. Finally, we assess the sensitivity of the total risk to each scenario. The optimal measure differs spatially throughout Afghanistan, but in most districts it is more beneficial to take flood DRR measures. However, in the districts where it is more beneficial to take earthquake measures, the reduction in risk is considerable (up to 40%, whilst flood DRR measures lead to a reduction in risk by 16% in individual districts). The introduction of asynergies between DRR measures in risk analyses allows policy‐makers to spatially differentiate building codes and other building‐level DRR measures to address the most prevalent risk while not compromising the risk resulting from other hazards.