Sub-tropical and tropical horticultural tree crops are known to have issues of low productivity. Understanding how and why canopy structural growth habits impose productivity limitations, is critical to managing high yielding, regular bearing mango (Mangifera indica) orchards. To investigate the relationship between shoot architecture, flowering, and fruiting in mango in Far North Queensland, we measured spatial and temporal patterns of branching and flowering at the growth unit level, under several varieties and training systems to see how tree size (height, trunk diameter, canopy dimensions and volume) and vegetative and reproductive growth responds to canopy pruning and management. This paper will outline the architecture data collection technique used in this study and will summarise the first two years of vegetative growth response of three mango varieties (Keitt, Calypso™ and NMBP1243) planted at three densities (208 trees ha-1, 450 trees ha-1 and 1250 trees ha-1) and trained as conventional or single leader. The minimum level of assessment is at the growth unit level (growth flush), because mango shoots develop from preformed buds which form an uninterrupted “growth unit” before terminating and reshooting. New growth units are produced two or three times per season followed by a quiescent or dormant stage. At the break of this dormancy, terminal growth units produce a terminal flower. While tree growth is generally seasonal, flowering depends on other endogenous and environmental factors including terminal growth unit age, bud functional history and strength of inductive signals. As a result, growth stages of the growth units can be asynchronous, often resulting in spatial and temporal divergence between vegetative and reproductive growth at different locations in the canopy, at the end of a growing season.