BACKGROUND For some phenomena the mean of population distributions predicts the proportion of people exceeding a threshold value. OBJECTIVE To investigate whether in depression, too, the population mean predicts the number of individuals at the extreme end of the distribution. METHODS We used data from the European Outcome in Depression International Network (ODIN) study from populations in Finland, Norway and the UK to create models that predicted the prevalence of depression based on the mean Beck Depression Inventory (BDI) score. The models were tested on data from Ireland and Spain. RESULTS Mean BDI score correlated well with the prevalence of depression determined by clinical interviews. A model based on the beta distribution best fitted the BDI distribution. Both models predicted the depression prevalence in Ireland and Spain fairly well. CONCLUSIONS The mean of a continuous population distribution of mood predicts the prevalence of depression. Characteristics of both individuals and populations determine depression rates.