The 170 samples obtained from ligneous investigation in Barkoissi, Galangashi and Oti-Keran protected areas (northern Togo) were subjected to floristic processing and several multivariate analyses in order to study plant diversity, life form and chorological type distribution and to classify and identify the main plant communities with their demographic structure. First, the overall life form and phytogeographical type frequencies were calculated to analyze and show the distribution of the 68 plants species in this area. Then plant communities were determined by using the hierarchical clustering with wards method. The plant communities found were identified by the technique of indicator value (IndVal). For each grouping diversity index such as specific richness, Shannon index and Pielous evenness was set out. The canonical correspondence analysis (CCA) was also used to relate the impact of environment factors and the distribution of samples. Finally the variance analysis (ANOVA) associated to post-hoc test was used to compare the similarity between the groupings. From these 170 samples, 12 plant communities were discriminated, three for the reserve of Barkoissi, three for the reserve of Galangashi and six for the reserve of Oti-Keran. The samples were well distributed in the factorial plan form by axe 1 and 3 of CCA. The plant communities were distributed along the moisture gradient in the ordination. The Sudano-Zambesian species followed by Sudanian species were the phytogeographical types most found in these areas. The microphanerophytes were the most represented life form. The structural parameters such as diameter and height means, the basal area and the densities determined and submitted to post-hoe test show some important similarity between some groups found. In overall, the vegetation condition of the three protected areas is somehow disturbed while most of the plant communities are stable. These preliminary ecological data obtained are essential and fundamental for the management and protection for these ecosystems.