1. The method of selection of a small group of schizophrenic subjects for long-term biological studies is an essential part of the research. In developing criteria the following concepts are considered: (a) range of variation of irrelevant variables, (b) random sampling, nonpurposive bias and purposive bias, (c) extension of the sample and intensification of pertinent characteristics of patients through the process of selection and evaluation.2. Variables presumbly irrelevant to schizophrenia are restricted: e.g., age, sex, race, length of hospitalization and complications of hospitalization and treatment. The acceptance of limitations inherent in a chronic schizophrenic sample presupposes that schizophrenia is a reflection of an on-going psychophysiological process (which however may appear only under certain operant conditions). The need for the study of acute schizophrenic and other samples is stressed.3. An operational framework for biological versus psychological factors in schizophrenia is considered; the assumptions involved are tabulated. The working hypothesis assumes that there is a group of schizophrenias. Within this framework, process schizophrenics, catatonic and hebephrenic subtypes and twins concordant or discordant as to schizophrenia are discussed as means for sample bias.4. A genetic-familial design is presented as the basis for a biased sample. It was decided that the widest sampling of genetic causes would be achieved if one-fourth of the sample had a family history suggesting dominance; one-fourth a history suggesting recessiveness, and one-half a negative family history. The patients with negative family history would be more likely to include an uncommon genetic type and also would render less likely the exclusion of the very group for which selection-bias is intended. The minimal conditions are as follows: (a) Family History Group–The Dominant Pattern; one schizophrenic parent; (b) Family History Group–The Recessive Pattern; neither parent affected but with at least one affected sibling; (c) Negative Family History Group; no history of schizophrenia within the family unit.5. The decreasing number of subjects available from a patient population of 13,500 in 3 collaborating hospitals is noted as the following additional criteria are imposed: diagnosis of schizophrenia, male, white, age range 18-40, duration of hospitalization of more than one year and less than 15 years, and familial pattern. A comparison of Family History and Negative Family History Groups with the initial criteria for selection is made; data referring to the first 14 admissions are presented.Abstract Teaser