This article describes two assumptions that are currently operative in biological research in schizophrenia. First, it is assumed that etiologic heterogeneity is present in this group of patients, even within patient cohorts diagnosed by precise, narrow research criteria. Secondly, it is assumed that overlap may exist with regard to specific biological abnormalities between schizophrenics and patients who satisfy research diagnostic criteria for other psychiatric disease entities. These assumptions suggest a need to relate putative biological abnormalities in schizophrenics to an array of historical, phenomenological, treatment response, long-term outcome, and family data in order to identify schizophrenic subtypes. Concern with biochemical abnormalities as “markers” for schizophrenia should be supplemented by efforts to demonstrate that the abnormality is relevant to the psychopathology of the perhaps small subgroup of patients in whom it is found. The following areas of research are reviewed for evidence of any usefulness in identifying schizophrenic subtypes: dopaminergic abnormalities; studies of other neurotransmitters or neuro-modulators, including peptides; neuroendocrine studies; platelet monoamine oxidase (MAO) activity studies; autoimmune phenomena; and HLA antigens. The importance of genetic studies for biochemical research in schizophrenia is emphasized.