Newly introduced rapid diagnostic tests for group A streptococcal pharyngitis should facilitate appropriate antimicrobial use in patients with group A streptococcal pharyngitis. Because of high rates of acute pharyngitis in Tuba City, AZ, at the Navajo Indian reservation, the use of rapid diagnostic test was prospectively evaluated. The sensitivity and specificity of the test was measured and changes in physician prescribing patterns attributable to use of the test were correlated. Of 320 patients with pharyngitis enrolled during the present 3-week study, 86 met the study's definition of a patient with streptococcal pharyngitis and 163 met the study's definition of a patient with nonstreptococcal pharyngitis. The rapid test was 62.8% sensitive and 96.9% specific in identifying patients from whom group A streptococci were isolated. Although treatment of patients with streptococcal pharyngitis at the time of the first visit increased from 36.5% in a retrospective sample to 72.5% during the study, treatment of patients in whom cultures were negative remained the same. Further analyses showed that physicians tended to treat patients with signs characteristic of streptococcal pharyngitis and, as the study progressed, to rely less on negative rapid test results as a reason to withhold antimicrobial agents. It was concluded that rapid tests with good specificity but limited sensitivity may improve treatment of patients with streptococcal pharyngitis by allowing earlier specific therapy. A more sensitive test with a higher negative predictive value would be necessary to prevent treatment of persons with nonstreptococcal pharyngitis.