Planting native trees in high diversity has been widely recommended to recover tropical forests to regain ecosystems that are taxonomically and functionally similar to the native forests. However, whether planting mixed stands would provide better results than planting pure stands remains to be demonstrated. Aiming to support seasonal tropical forest restoration in the highly fragmented landscape of the Atlantic Forest, Brazil, we compared the long-term (from 24 to 53 years) outcomes of pure and mixed plantings of native tree species with a broad spectrum of reference ecosystems (degraded, secondary, and old-growth forests). We aimed to verify if pure plantings would be as successful in fostering natural regeneration as mixed plantings and if they could potentially reach the attributes of native forests in the same region. We assessed forest structure (density, biomass, basal area), richness, and taxonomic and functional diversity. We separately analyzed the upper, intermediate, and lower layers of the tree community and assessed the abundance of ferns and climbers. Functional traits included seed dispersal and pollination syndromes, growth rate, seed mass, and specific leaf area. Contrary to our expectations, pure and mixed plantings did not differ in tree density, species richness, and functional diversity. Mixed plantings had higher basal area and biomass, surpassing the reference ecosystems and higher taxonomic diversity, whereas pure plantings had more ferns, climbers, and saplings regenerating in the understory. There was no difference for most of the functional traits between the forest types, except in seed mass, which was higher in mixed plantings. Zoochorous species were dominant in all forest types and size classes. In pure plantings, fast-growing species prevailed, whereas moderate and slow-growth species predominated in the other forest types. Pure plantings did not differ from the reference ecosystems, except in taxonomic diversity and richness when compared with the old-growth forests, which takes longer to be achieved. Our results indicate that the sparse remnants of forests are still effective sources of propagules to colonize the plantings – whether using a single species or a mix – even in a highly fragmented landscape, which was the case in our study. In general, pure plantings were as effective as mixed plantings to foster the recovery of seasonal tropical forests in terms of structure and functionality in the long term. As for taxonomic diversity and richness, mixed plantings were more efficient, whereas pure plantings have not reached old-growth forests but did not differ from degraded and surpassed secondary forests in several attributes. These results highlight the relevance of planting trees, even in low diversity, for increasing biodiversity and providing ecosystem services on a landscape scale.