Crop and mallee (Eucalyptus kochii subsp. plenissima (C.A.Gardner) Brooker and Eucalyptus horistes L.A.S. Johnson & K.D. Hill) growth and water use were measured in an alley system from 1999 to 2003. The aims of this study were; to quantify the growth and water use of agricultural crops and mallee belts grown as a short rotation woody crop and to test the hypotheses that a mallee agroforestry system is more productive than annual crop or mallee monocultures, and that managing competition in a mallee agroforestry system by root pruning can increase the productivity of the annual crop component. Mallee growth was typical of values reported for the Western Australian wheatbelt. Root pruning or harvesting mallees reduced mallee growth and water-use and partitioned more water to annual crops, but didn't increase the overall productivity of the system. As the mallees had exhausted stored soil water and didn't have access to fresh groundwater there was little complementarity in resource use. Rather than planting mallees in alley configurations (integration) there is a case for block planting (segregation) with longer harvest intervals for the mallees. The results suggest that biological productivity of mallee monocultures harvested on a rotation of five or more years > crop monoculture > mallee agroforestry. Ultimately the relative economic returns from cropping and mallees will determine where they are grown and the degree of integration. (C) 2015 Elsevier Ltd. All rights reserved.