Potential gains in crop water productivity of maize and wheat: Realistic gains or methodologically induced noise?

20–40% gains in Crop Water Productivity (CWP) have been widely advocated in the literature. These claims, however, are difficult to square with the physical relation that exists between transpiration and biomass accumulation. We reviewed and re-analysed published field studies on CWP of wheat and maize conducted between 1983 and 2023 to assess the rigour and accuracy of these results and claims. Our results show that 90 % of the publications do not comply with a rigorous methodological assessment of actual evapotranspiration by solving the water balance, leading to frequent and structural over- or under-estimations of CWP. Only 4 per cent of publications met the methodological criteria we set for ET a assessment and accounted for the moisture content of the reported yields. This drastically reduced the available datasets (3–5 studies per crop) to draw conclusions on CWP relations and potential gains. It also drastically reduced the range of CWP values recorded, implying potential gains in CWP may be much more limited than claimed. Based on the delimited available empirical basis, CWP improvement practices (deficit irrigation, mulch and nutrient availability) were assessed through disaggregated CWP analysis and show gains in CWP are feasible but delimited in scope. We end with an urgent call to journal editors and fellow scientist to improve the rigour of our methodological assessments of CWP, providing a minimum set of criteria to be met, and enlarge the empirical base of good quality and uniform data from which we can set benchmarks for crop specific CWP improvements in the future, that are both realistic and attainable.