The effects of two putative attention-engaging maneuvers on tracking performance were studied in three groups of subjects: inpatients (n = 19), outpatients (n = 19), and controls (n = 20). One method involved realerting subjects during tracking by repeating instructions to track carefully. The second method, a signal detection task incorporated into the tracking stimulus, required that subjects signal their perception of brief interruptions of the tracking light by pressing a hand-held button. The tracking performance of inpatients was significantly inferior to that of both outpatients and controls, whereas tracking performance of these two latter groups did not differ. Verbal realerting did not significantly improve tracking performance in any group; moreover, during the administration of these instructions there was an increase in tracking errors in inpatients. Inpatients also made more tracking errors than comparison groups during signal detection trials. Other subject factors of possible relevance to tracking performance, e.g., age, gender, and level of arousal, were found not to covary with tracking accuracy in a manner which would explain the observed group differences. It is unlikely that voluntary inattention is the basis for the observed impaired tracking in hospitalized psychiatric patients; the data are more consistent with an interpretation based on heightened distractibility or information overload in these patients.