While it may be feasible to target those receiving relief after a natural disaster, it is difficult to differentiate the amount of relief provided among beneficiaries. This is because much of the relief consists of food, clothing, and medicine, all goods for which the absorptive capacity of households is limited. Empirical tests using data from Honduras following Hurricane Mitch confirm this hypothesis. The probability of receiving relief was negatively correlated with wealth and positively correlated with assets losses (with a higher weight placed on losses than pre-disaster wealth) and the fact that households suffered damage to their dwelling. By contrast, controlling for whether households suffered damage to their dwelling, the amount of relief received was related neither to pre-Mitch wealth, nor to assets losses