Clinical utility of completion time and performance errors was investigated for the Trail Making Test (TMT; Reitan, 1958). Archival neuropsychological files for patients with mild and moderate/severe head injuries, as well as patients with suspect effort on neuropsychological testing, were examined and compared to controls and experimental malingerers. Time-to-completion scores differentiated the participants who were malingering and who gave suspect effort from those patients with head injuries. There were no differences in error rates among the head-injury groups or controls for either the TMT-A or TMT-B. Errors were also not uncommon among normal controls; 12% and 35% of the controls made at least one error on TMT-A and TMT-B, respectively. However, error rates for both the suspect-effort and malingering groups were inflated on TMT-B as compared to the head-injured and control groups. Results suggest that performance errors on the TMT lack diagnostic utility for persons with head injuries, and time-to-completion is still the best indicator of neuropathology. However, performance errors, in conjunction with inflated time scores on the TMT, may be useful in the assessment of malingering.