In Fennoscandia, the species richness of vascular plants in 75 × 75 km squares is highly correlated with geographical (latitude and longitude) and climatic variables (accumulated respiration sum, mean January temperature, and mean July temperature). When generalised additive models (GAM) are used, over 80% of the variation in richness can be statistically explained by geography and climate. Even though climate has such a high explanatory power we present several arguments for interpreting these results with care. Climate has no ecologically sound explanatory power when the variation due to latitude and longitude is accounted for, and the strongest latitudinal gradient in summer temperature is in an area where the latitudinal gradient in species richness is absent. We discuss the role that Holocene history might have on the variation in species richness, and argue that history and climate should be considered simultaneously when explaining the observed patterns in the geographical variation of species richness.