Ecoregional research has the potential to help address some of the huge challenges facing agriculture in developing countries by developing technologies that work under different agro-ecological conditions, and the processes by which these technologies can be adapted to work in other areas with similar conditions. The CGIAR system has been developing ecoregional research as a new paradigm for over a decade. In this paper we evaluate one of the most ambitious of these initiatives called the Benchmark Area Approach (BAA) pioneered by the International Institute of Tropical Agriculture. We evaluate the BAA against nine good practice criteria for ecoregional research and finding that the approach is delivering, or has the potential to deliver, on all nine. Many of the lessons learnt from this evaluation will be relevant to current and future attempts to undertake co-ordinated multi-locational research for development.