Developing objective tools for tracking progress of restored sites is of general concern. Here, we present an innovative approach based on principal response curves (PRC) and species classification according to their preferential habitats to monitor changes in community composition. Following large-scale restoration of a cut-over peatland, vegetation was surveyed biannually over 8 years. We evaluated whether the establishing plant communities fell within the range of natural variation. We used both general diversity curves and PRC applied on plant species grouped by preferred habitat to compare restored sites and unrestored sites to a reference ecosystem. After 8 years, diversity and richness differed between the sites, with Forest and Ruderal species more prominent in unrestored sites, and Peatland, Forest, and Wetland species dominant in restored sites. The PRC revealed that the restored site became rapidly dominated by typical peatland plants, the main drivers of temporal changes being Sphagnum rubellum, Pohlia nutans, and Mylia anomala. Some differences remained between the restored and the undisturbed species pools: the former had more herbaceous species associated with wetlands such as Calamagrostis canadensis and Typha latifolia and the latter had more forested species like Kalmia angustifolia throughout the study. PRC revealed to be an efficient tool identifying species driving changes at the community level after restoration. In our case study, examining PRC scores after classifying species according to their preferred habitat allowed to illustrate objectively how restoration promotes target species (associated to peatlands) and how lack of intervention benefits ruderal species.