This study is one of few detailed analyses of plant diversity and vegetation patterns in African dry forests. We established permanent plots to characterize plant diversity, above-ground biomass, and vegetation patterns in a tropical dry forest in Kimbi-Fungom National Park, Cameroon. Our results contribute to long-term monitoring, predictions, and management of dry forest ecosystems, which are often vulnerable to anthropogenic pressures.Considerable consensus exists regarding the importance of dry forests in species diversity and carbon storage; however, the relationship between dry forest tree species composition, species richness, and carbon stock is not well established. Also, simple baseline data on plant diversity are scarce for many dry forest ecosystems. This study seeks to characterize floristic diversity, vegetation patterns, and tree diversity in permanent plots in a tropical dry forest in Northwestern Cameroon (Kimbi-Fungom National Park) for the first time.We studied associations between above-ground biomass and species composition, and how different vegetation types vary in terms of species composition, diversity, and carbon storage, in a dry forest in Kimbi-Fungom National Park, Cameroon. Vegetation was inventoried in 17 permanent 1-ha plots. Allometric equations were used to calculate above-ground biomass and carbon.We found an average of 269.8 tree stems ha and 43.1 species ha. Five vegetation types: semi-deciduous, gallery, mixed vegetation, secondary and the grassland/woody savanna forest were classified using TWINSPAN analysis. The five vegetation types had an average above-ground biomass of 149.2 t ha and 74.6 tC ha of carbon in the 17 ha analyzed. Canonical correspondence analysis (CCA) showed the importance of semi-deciduous forest over grassland/woody savanna forest.This study demonstrated that the forest of the Kimbi-Fungom National Park is poor in plant diversity, biomass, and carbon, highlighting the need to implement efficient management practices. Fine-scale inventory data of species obtained in this study could be useful in developing predictive models for efficient management of tropical dry forests.