A grip-induced muscle tension maintenance task distinguished between schizophrenic patients, regardless of medication or hospitalization status, and both normals and controls with affective (unipolar and bipolar) disorders. Unaffected first-degree relatives of schizophrenic patients also showed a grip deficit. Coupling the grip task with a visual discrimination task that divided attention through instruction or reinforcement contingency, increased grip error times for all groups. No group differences appeared on the discrimination task, regardless of difficulty level, and the tone used to provide corrective feedback was not implicated in the grip deficit. The results suggest that the grip task is tapping, in a systematic and reliable manner, a motor-control abnormality that may be useful as a behavioral marker of schizophrenia.