Five hundred twenty-five patients were selected according to specific research criteria from records of 25 to 35 years ago. Diagnoses appended to charts at that time are compared with research diagnoses. Of patients originally diagnosed as having affective disorder, 25% were discarded from research sample. Of patients originally diagnosed as schizophrenics, 63% were not included in research sample, mainly because of acute onset or episodic course. A high degree of correlation between clinical and research criteria for affective disorders was found, but criterion of "agreement amongst experienced clinicians" relatively poorly satisfies research criteria for schizophrenia.

Very stringent research diagnostic criteria were used; but for 200 schizophrenics diagnosed systematically, there was chart agreement for 95%; and in 325 affective disorders diagnosed systematically, there was 93% agreement with original chart diagnoses made in 1935.