OBJECTIVE--To develop and evaluate a short (10 item) simple measure of outcome mainly for use with patients with mild to moderate head injuries. DESIGN--Two studies on patients at three and six months after injury, comparing different methods of administration (two raters and postal questionnaire), and comparing ratings with other assessments. SUBJECTS--Forty three patients seen three months after injury and 46 seen six months after injury; both groups had head injuries covering a range of severity from minor to severe. MAIN OUTCOME MEASURES--Differences between ratings in different groups of patients (Mann-Whitney U test); differences in ratings used different methods of administration (Wilcoxon signed rank test); and correlation between ratings from the same patient (Spearman r). RESULTS--The sum total ratings were consistent between raters and between methods (postal questionnaire v face to face interview) with no evidence of selective bias between rates or methods. Ratings on individual items were also reasonably consistent. The sum total rating varied as anticipated between groups divided by clinical judgement of recovery and patient assessment of recovery, and related as expected to the extent of post-concussion symptomatology. The 10 items included covered the most important problem areas reported by patients. CONCLUSION--The Rivermead head injury follow up questionnaire (RHFUQ) is a short, simple, adequately reliable, and valid measure of outcome, across the entire range of severity, but particularly after mild to moderate head injury.