This study proposes a preference-based benchmarking approach to assess trade-offs between alternative uses of aid dollars. We ask ~800 low-income Kenyans their valuation (in cash) for common aid and development programs. We compare recipients’ stated valuations to the cost of each program and to the valuations of a population working in the development industry. We find that recipients value some common programs less than the cost of delivery. On an absolute basis, development professionals also value certain interventions less than the cost of provision, but not always the same interventions as recipients. While development professionals and recipients are in accord on the raking of value for cost according to broad categories (e.g., public vs. private goods), they place different relative value weights on specific interventions. Thus, in a world with limited resources a portfolio of interventions selected by development professionals could be significantly less valuable in the eyes of recipients than recipients’ preferred allocation.