Rather than using more or less ideal conditions for setting experimental controls, the use of conditions similar to those likely to be encountered by farmers should produce research results which are realistically achievable on the farm. ICLARM has developed an approach to farmer-led experimentation which utilizes a spreadsheet to collate and analyze data collected from participating farmers. The simulation of actual management practices utilized by farmers produced results in replicated on-station trials which were within 11% of net yields on-farm. In addition to giving researchers a tool for comparing farm and station management practices, giving farmers a realistic indication of what yields will be if a technology is adopted should help overcome the problems of disillusionment often encountered when farm results fall below those expected by researchers on the basis of experiment station studies.