Aim: Understanding the drivers of global diversity has challenged ecologists for decades. Drivers related to the environment, productivity and heterogeneity are considered primary factors, whereas disturbance has received less attention. Given that fire is a global factor that has been affecting many regions around the world over geological time scales, we hypothesize that the fire regime should explain a significant proportion of global coarse-scale plant diversity. Location: All terrestrial ecosystems, excluding Antarctica. Time period: Data collected throughout the late 20th and early 21st century. Taxa: Seed plants (= spermatophytes = phanerogamae). Methods: We used available global plant diversity information at the ecoregion scale and compiled productivity, heterogeneity and fire information for each ecoregion using 15 years of remotely sensed data. We regressed plant diversity against environmental variables; thereafter, we tested whether fire activity still explained a significant proportion of the variance. Results: Ecoregional plant diversity was positively related to both productivity (R2 = .30) and fire activity (R2 = .38). Once productivity and other environmental variables were in the model (R2 = .50), fire regime still explained a significant proportion of the variability in plant diversity (overall model, R2 = .71). The results suggest that fire drives temporal and spatial variability in many ecosystems, providing opportunities for a diversity of plants. Main conclusions: Fire regime is a primary factor explaining plant diversity around the globe, even after accounting for productivity. Fires delay competitive exclusion, increase landscape heterogeneity and generate new niches; thus, they provide opportunities for a large variety of species. Consequently, fire regime should be considered in order to understand global ecosystem distribution and diversity.