“Study objective - The aim was to examine why differences exist in schizophrenia prevalence and risk in some areas of Croatia, when schizophrenia incidence rates do not appear to vary. Design - Areas differing by schizophrenia admission rates in patients born in 1953 and admitted by the age of 31 years are compared using a number of indicators relating both to general population characteristics and to those of schizophrenic cases in these populations. Setting - The study covers the whole of Croatia (4 601 469 inhabitants, 1981 census). Subjects - By the age of 31 years, out of 80 445 individuals born in Croatia in 1953, 464 were admitted for and diagnosed as having schizophrenia. Main results - Admission risk rates are higher in those parts of Croatia where emigration rates are high and lower where immigration rates are high. There is also a positive correlation with schizophrenia prevalence and manic depressive psychosis rates. There is a negative correlation with age of onset of schizophrenia and with schizophrenic reproduction rates. In the study areas, hospital incidence rates are not significantly different. Conclusions - Economic migration and negative selection in the domestic population are likely to be the most significant factors leading to differences in schizophrenia prevalence. The approximately equal incidence rates in the population, with different prevalence and admission risks, are linked to differences in the disease onset among schizophrenics with a positive family history for this condition. In other words, these patients, when part of the population with a greater prevalence and a greater hereditary loading, experience the onset more often at an earlier age. Thus they have a lower reproduction rate than in a population with a lower prevalence and a lower hereditary loading. Thus incidence rates in populations with different prevalences and different hereditary loads are maintained roughly equal over generations.”