Loosening of thinking as assessed by the Object Sorting Test (OST) has been found in a percentage of normal subjects but in a higher percentage of schizophrenics, and is familially transmitted in both groups. Loosening of thinking in normal subjects is not associated with evidence of impaired function or increased psychopathology, and in recognition of this, it was termed allusive thinking rather than thought disorder. Both OST-assessed loosening and concreteness of thinking were found to be present independently in a high percentage of schizophrenics, so that both were considered to contribute to schizophrenic thought disorder. The presence of OST-assessed loosening in schizophrenics would, therefore, be predicted to correlate partially rather than totally with measures of schizophrenic thought disorder. It has been suggested that OST-assessed loosening in normal subjects is due to a genetically determined reduction in strength of an inhibitory process that limits the spread of activation of semantic associations and results in a predisposition to schizophrenia. The brain event-related potential P300, which is, in part, under genetic control, may index this inhibitory process. Therefore, it was predicted that in normal subjects, P300 would correlate with OST-assessed loosening of associations. If schizophrenic thought disorder is due to a further weakening of this inhibitory process, it can be predicted that P300 in schizophrenics correlates only weakly with OST-assessed loosening of thinking, but more strongly with schizophrenic thought disorder. In a study in which P300 was elicited using a difficult selective attention task with 15 unmedicated schizophrenics and 22 healthy subjects, all three predictions were supported.