In 153 male schizophrenic probands born in 1881–1900 the age range of onset ran from 9 to 77 years, the mean and standard deviation were 35.6^12.6 years. By securing probands from a sufficiently distant birth period one avoids the bias caused by dilution of the material with index patients with early manifestation, who get a disproportionally high chance of ascertainment by sequential admission sampling. By calendar year of birth 748 full sibs of probands ranged from 1865 to 1919, with an age corrected incidence of schizophrenia with hospital care of 11.6%. A mean number of 3 sibs per proband were still alive when 50 years old. The birth period approach reduced losses of sib material through curtailing the observation period to less than 0.2%. Together with varieties within sibships the extended risk period for schizophrenia adduces suggestive evidence against autosomal recessive inheritance as a major cause of this psychosis.