Questions: How many vegetation classes and alliances are described in European countries? What are the main factors driving the number of these vegetation types? What would be the estimated number of vegetation types in so far under-explored regions outside Europe? Location: Twenty-three countries and regions in Europe. Fourteen regions outside Europe. Methods: We compiled lists of vegetation types (phytosociological classes as broader units and alliances as narrower ones) described in European countries with a well-developed vegetation classification. The delimitation and nomenclature of classes and alliances were standardized according to a unified classification system for Europe. Generalized linear models were used to test the influence of floristic richness, area, sampling effort, altitudinal range and geological diversity on the number of vegetation types. Best predictors were used to estimate the number of classes and alliances in selected regions over the world. Results: In agreement with broadly recognized diversity patterns, we found the highest numbers of vegetation types in countries with a mediterranean climate. The numbers of classes and alliances per country did not depend on the country size, while country's floristic richness was identified as the main explanatory variable, accounting for up to 67% of explained deviance. The number of alliances increased in countries with broad altitudinal range and large geological diversity, but these variables did not influence the number of classes. Accordingly, estimates of the number of alliances for other (mostly non-European) regions were adjusted when including altitudinal range as an additional predictor. Conclusions: At broad scales, the number of vegetation types can be predicted from floristic richness, which indirectly accounts for the effect of area, and from environmental heterogeneity. Furthermore, this overview demonstrates that the number of phytosociological alliances reflects biogeographic patterns in Europe better than the number of classes. We also estimate, for the first time, the number of vegetation types that may be expected in so far poorly surveyed regions worldwide, using procedures of vegetation classification analogous to those used in European phytosociology. © 2013 International Association for Vegetation Science.