A study was conducted to evaluate the usefulness of brush cytology as a diagnostic tool for lesions of the head and neck. Brush specimens were obtained from patients for whom surgical biopsy was recommended by the Otolaryngology Service of the University of Texas Medical Branch Hospitals, Galveston. Specimens were interpreted independently by three cytologists, and interobserver variability was calculated. If a surgical biopsy was performed, histopathologic diagnosis was used as the reference standard to assess the accuracy of each cytologic interpretation. Correspondence of cytopathologic interpretation with histologic diagnosis was sufficient to conclude that brush biopsy is a useful screening technique for unsuspected or clinically undetected malignant tumors of the upper aerodigestive tract. Because of its ability to sample large surface areas with minimal tissue trauma, brush biopsy can be a useful screening technique in combination with selective surgical biopsy for the detection of cytologic changes of malignant neoplasia. Brush cytology costs less than surgical biopsy, yet its relatively high sensitivity and specificity for both benign and malignant grades support its utility. Brush cytology, furthermore, has a low interobserver variability for the benign and malignant grades, suggesting that in the hands of an experienced cytopathologist it can be relied on with confidence. For grades 2 or 3 (inconclusive), brush cytology, however, demonstrated much higher interobserver variability. Based on the findings of this and other studies, brush cytology can be effective in identifying clinically unsuspected malignant tumors of the upper aerodigestive tract, especially in patients with mucosal changes suggestive of "field cancerization."