The incidence of paranoid schizophrenia (the number of new cases per year) was examined in a population. The figure was found to be 0.17 per 10,000 of the population, being greatly dependent on the age and sex structure of the population examined. The method of the examination is presented in detail. It is shown, in particular, that the results are considerably influenced by the disease duration in the patients. With a short duration of the disease the results appear to be greatly underestimated because of the difficulties in making the correct diagnosis in many cases. The exact number of such cases can be determined by using the law of statistical distribution of the duration of the paranoid schizophrenia stages.