Three modifications were made to the Sickness Impact Profile, a behavior-based measure of health status, to improve its sensitivity to the effects of head injury. (1) Additional items were included to capture head injury sequelae and behaviors typical of young adults, the age group to which head injury most frequently occurs. (2) Subjects individually excluded behaviors irrelevant to them, thus allowing the score to better reflect injury-related changes. (3) The different areas of functioning on the Sickness Impact Profile were reweighted to reflect global judgments of the construct's contribution to overall functioning rather than the sum of the item contributions. Only the first modification is head-injury specific. The others, are relevant to any disease or injury. The performance of the modifications was evaluated in a longitudinal study of 102 head injured and 102 comparison subjects tested at 1 and 12 months after injury. The evaluation of the modifications was based on their ability to distinguish head injury from comparison subjects and on the strength of their relationship with measures of brain dysfunction. Despite a few statistically significant improvements in discrimination, differences of a practical degree were not obtained. The standard Sickness Impact Profile performed well and is recommended for evaluation of day-to-day functioning in head injury studies.