Abstract Understanding co-adapted traits and strategies in organisms has been increasingly shown to require a broader view of structure, function, taxonomic affiliation (phylogeny) and their possible interactions. In order to understand mechanisms of recruitment dynamics in tropical forest, nine juvenile (fruit, seed and seedling) and seven adult ecological and morphological traits were examined in 89 woody plant species, from 75 genera, 35 families and 21 orders. The species were from tropical rainforest in northern Queensland and varied widely in light requirements and hence regeneration niches. The importance of taxonomic (phylogenetic) constraints was considered using various techniques. Data were collected mostly from herbarium specimens, augmented by observations in the field. The main source of variation in most traits was at the species level, although fruit size and light requirement varied most at the genus level, and flower size at the level of order. Covariation among attributes increased in strength with increasing taxonomic level. Cluster analyses of the species arising either from five or three groups reflected juvenile rather than adult attributes. However, these groups were not morphologically distinct. Sets of juvenile traits poorly predict adult functions (and vice-versa) at species, genus and family levels, but do so more strongly at the order level. Analyses of variance and multivariate analyses, with and without the effects of phylogeny, indicated that fruit size, seed size, hardness of seed coat, seedling cotyledon function, fruit maturation period and, to a lesser extent, seed number per fruit and wood density could predict light requirement, and hence regeneration niche. Flower size showed a strong phylogenetic effect, and thus is conserved along taxonomic lines. It is concluded that as different traits may vary at different taxonomic levels, the influence of phylogeny on interspecific variation needs to be routinely recognized by ecologists involved in comparative analyses of plant life-history functions.