Productivity in temperate tree crops such as apple has been lifted several-fold by research focusing on optimising a combination of canopy components including light relations, vigour control, tree architecture and crop load. This paper outlines the research behind the Small Tree-High Productivity Initiative (STHPI), which is focused on improving productivity of mango, avocado and macadamia. Preliminary results from work we are undertaking for each of the above canopy components in mango will be outlined. A rootstock screening trial to identify vigour-managing, high-productivity rootstocks is being undertaken, and we present a comparison of the best-performing low-medium vigour rootstock varieties compared with control 'Kensington Pride' (KP) rootstock at 6 months old. Comparisons between 'Keitt', 'NMBP 1243' and 'Calypso' scion cultivars with regard to tree diameter, height and canopy growth at different orchard densities and training systems will also be presented. Preliminary results from an orchard light-relations study indicate that mango yields continued to increase with light interception up to 50%, and reached a maximum of 20-30 t ha-1 at 68% light interception in KP trees approximately 25 years old. In a crop load trial, inflorescence thinning in a 'Calypso' orchard did not significantly reduce yields when up to 90% of inflorescences were removed, but did when 95% of inflorescences were removed, as trees were unable to compensate by adjusting fruit set, size and yield. Inflorescence thinning beyond 80% increased the number of fruit set per panicle, and thinning up to and including 90% of inflorescences increased fruit weight from 340 g to >400 g per fruit. This project is still in its initial stages; however, early indications suggest there may be opportunities to improve early orchard yields through optimising light interception in an orchard's life, potentially through the use of higher densities, and that rootstocks and tree training methods, once identified, may help in the management of vigour. It is also hoped to obtain a better understanding of how crop load influences the balance between vegetative growth, flowering, fruiting, alternate bearing and fruit quality.