The proportion of "true" cases of schizophrenia is high in the Swedish National Inpatient Register. In addition, the results did not support large differences between clinics or regions although a larger sample size would be needed to test this formally. In comparison with earlier studies from the Nordic countries the validity seems to be quite stable over time, between different ages at onset and between Nordic countries. Thus, the registers are well suited to be used in schizophrenia research.