Measuring impacts of non-native species is challenging, particularly for long-established invasions in poorly characterised native ecosystems. Recreating historical population dynamics using dendrochronological methods is a popular approach for addressing conservation and forest management questions. However, these tools are rarely applied for addressing questions of invasion. Classical dendrochronology is also not well suited to many tropical tree species as they do not produce growth rings. Here we test the validity of an alternative approach, based on the analysis of multiple morphological markers, that could allow for both the recreation of historical invasion dynamics as well as quantify the impacts of invasion on the recipient native ecosystem. We validate this method on the Cecropia peltata invasion in Cameroon and its potential impacts on a native tree, Musanga cecropioides. We were able to compare architectural development over time in a mixed secondary forest where the two species co-occur and revealed differences in performance (growth, branching and flower production) between the two tree species. Our results revealed more sustained growth during early establishment and an early sexual maturity for the non-native C. peltata, compared to its native counterpart. A lower density of M. cecropioides was observed in the region of co-occurrence, supporting a hypothesis of competitive exclusion of the native species at the juvenile stage. Our work provides a novel way to understand the impacts of plant invasions via retrospective analysis that could well be applied to other invasions globally, opening up more opportunities to prioritise management on an impact-focused basis.