Previous studies have indicated that patients with mild head injury have no specific deficits in selective attention as assessed with the Stroop Color Word Test. It has been reported that patients may be slower to name colors, but that they are not disproportionately slower on the subtask with color-word interference. The present study examined whether the performance of patients with mild head injury (MHI) was worse than that of nonconcussed controls on a modified Stroop interference subtask of increased complexity. Patients (n=44) were individually matched with controls for age, sex, and education. It was found that MHI patients were significantly slower on this modified subtask, whereas there was no significant difference between the two groups with the original interference subtask. The findings support the hypothesis that the cognitive disturbances of patients with mild head injury may be subtle and only demonstrable with very sensitive tests.