Sixty children (120 ears) were evaluated using reflectometry, pure tones, immittance, otologic examination, and pure-tone air and bone conduction thresholds. Comparisons were made between the various screeners. Reflectometry was found to be the least sensitive and the least specific of the procedures. Results were then examined within each diagnostic category to determine test agreement with the otologists' judgments. The reflectometer categorization was found to be inconsistent with the otologic findings. Finally, the reflectometer was compared to each of the other screener results and pure-tone thresholds. The test results were significantly correlated but showed high over- and under-referral rates for the reflectometer. Seventy-eight children and adults (156 ears) were screened in a second study. Again, the acoustic reflectometer did not perform adequately as a general screening tool. When only cases of middle ear effusion were screened, the acoustic reflectometer did not perform as well as immitance.