Question: How far can we simplify the floristic complexity of a tropical rainforest into functional groups in order to predict tree population dynamics after logging-induced disturbance? Location: Paracou experimental site, French Guiana. Methods: We used data from over 15 years in control and disturbed plots from a silvicultural trial started in 1984. We selected 53 common tree species assigned to five functional groups based on potential size and light requirement. For each species, we quantified: the fate, i.e. variation in population size, and dynamic processes, i.e. mortality, recruitment and growth, driving this fate. We investigated the links between dynamic processes, fate and functional groups. Results: Disturbance stimulated growth and recruitment for most species, but had a heterogeneous impact on mortality. Species fate in disturbed plots depended on recruitment and was more favourable than in control plots. The functional classification was more predictive for most separate dynamic processes than for species fate: after disturbance, significant differences were found between all functional groups for growth. Pioneer+heliophilous species showed significantly higher recruitment rates. Mortality of shade-tolerant species slightly increased and of mid-tolerant and heliophilous species decreased. Conclusions: A combination of three species classifications separately built from the growth, recruitment and mortality processes is more informative than a global classification combining the processes. Identifying the pioneer+heliophilous species on the basis of their growth rate is crucial to predict species fate after disturbance. We showed that potential growth rate could be used as a reliable indicator to identify this group.