HIS PAPER is concerned with the make-up of the urban population in mental hospitals: What kinds of persons the patients are, what types of urban areas they come from, and what kinds of social life characterize those areas. Special attention is paid to social isolation as a possible cause of schizophrenia. From a large scale study of Mental Disorders in Chicago, by the writer and H. W. Dunham, there emerge some very clear answers which may be typical of urban patients in general. Some parts of this study have been repeated in Providence, Rhode Island, with identical results. The details, which may be published soon cannot be presented here, but some of the principal findings can be summarized. A most significant general finding is that different answers to the above questions must be given for the different types of mental disorders. In answer to the first question concerning what types of persons constitute the hospital population, each index measured shows great differences for the different diagnoses. Omitting details, the true picture can be indicated by certain index figures. Age, for example, varies greatly in the different diagnostic groups. The peak age group for the senile psychosis, that is, the age group with the highest rate, is 75 years and over. The peak age group for general paralysis, manic-depressive, and alcoholic psychosis is 45 to 54 years. The peak age for paranoid schizophrenia is 35 to 44 years, and for hebephrenic and catatonic schizophrenia, 25 to 34. The range of ages as well as the peak age group also varies with each diagnostic group. Marital status shows similarly great contrasts. In the same sample of cases, the ratio of married to single, taking single as ioo, is, in the general paralysis group, 3.I9, in the senile group 2.25, in the alcoholic group 2.07, in the manic-depressive group i.82, and in the schizophrenic group .64. Some differences of marital status also appear between the manic and depressed types, between the types of schizophrenia, and in each of the schizophrenic types between the sexes. Though no direct measure of economic status of these patients is available, it is possible to construct a serviceable index by using the median rentals in the areas from which the patients come. This procedure yields a symbolic monthly rental figure of $62 for the manic-depressive, of $39 for general paralysis, of $37 for paranoid schizophrenia, of $35 for senile psychosis, of $30 for hebephrenic schizophrenia, of $29 for catatonic schizophrenia, and of $23 for alcoholic psychosis. This represents a great spread