It is critical to understand how forests regenerate after the exclusion of human induced disturbances because the regenerating species drive the renewal of resources and ecosystems, which in turn support human beings locally and globally. This study of forest regeneration was conducted in the tropical coastal forest ecosystems of Tanzania at Uzigua Forest Reserve (24,730 ha) in the Pwani Region. We collected data from adult trees in 47 sampling plots (25 9 25 m) randomly established in closed forest sites (control) and in sites disturbed by farming and livestock grazing. Sapling and seedling data were collected and analyzed in 2 9 2 m nested subplots. Plots in the closed forest had higher mean basal area and volume for adult trees than those in sites disturbed by farming and livestock grazing. Plots in sites disturbed by farming had the highest Shannon–Wiener index for seedlings and saplings, followed by closed forest and livestock grazed sites. Closed plots and farmed plots had higher Simpson’s index for seedlings and saplings than in grazed sites. Plots in farmed and livestock-grazed sites had a higher Simpson index for adult trees than in closed sites. Equitability of seedlings and saplings was highest in closed forest sites, followed by farmed and livestock grazed sites. Plots disturbed by farming had higher equitability for adult trees than in livestock-grazed and closed sites. Moreover, plots disturbed by farming and grazing had a higher importance value index for seedlings than in closed sites. Plots in closed forests had higher sapling importance values than in farmed and grazed sites. Again, plots in closed forest had a higher importance value for adult trees than in farmed and grazed sites. The differences in tree species subcategories of regeneration across land-use indicates that farming and livestock grazing disturbances create heterogeneous microhabitats, which positively or negatively affects regeneration capacity of species after exclusion.