The variability of protection rates within sectors is frequently particularly high in agriculture relative to non-agriculture. Standard aggregation procedures ignore the variability within sectors, and underweight the importance of highly protected sectors. It therefore seems likely that they underestimate the potential benefits of agricultural trade reform relative to non-agricultural reform. This study examines this question using a new procedure for aggregating trade distortions. It finds that the key impact of using better aggregators is to increase the benefits of both agricultural and non-agricultural reform. It finds that using optimal aggregation procedures increases the measured importance of agricultural trade reform relative to non-agricultural reform from a very high initial level, but only by around two percentage points.