Farmers’ increasing abilities to make multiple measurements and to undertake quanti- tative comparisons create the potential not only to improve their on-farm decision-ma- king, but through shared ‘action research’, to improve the precision and sophistication of crop science generally. Thus, a Yield Enhancement Network (YEN) was set up in the UK in 2012 whereby farmers and scientists adopted metrics to explain each crop’s per- formance in terms of resource capture and conversion as well as in terms of agronomic actions. Wheat yields associated more with total biomass than harvest index, and more with grains m-2 than weight grain-1. Concomitantly, yield associated more with the fre- quency than with the quantity of boughten inputs and differences arose more between farms than through any other factor. In conclusion, the means favoured for yield en- hancement and the routes to achieve it must change. Additionally, agronomic variation must be understood through synthesis of metrics across multiple sciences, including those of farming behaviours.