PURPOSE: to describe the results of three TEOE protocols obtained from automatic equipment used in a Newborn Hearing Screening Program. METHOD: during two months, TEOEs of 287 neonates in São Paulo were studied. To register them, we used a three-protocol, set up with different pass/fail criteria. The following protocols were registered: Protocol A: to pass in four frequency bands, not necessarily consecutive-; Protocol B: to pass in three frequency bands, not necessarily consecutive - and Protocol C: to pass in two frequency bands, not necessarily consecutive, as well. The parameters that we used to consider these answers were: reproducibility above 50%, as well as signal/noise ratio > 3dB at 1.0 and 1.5 kHz and > 6dB at 2.0, 3.0 and 4.0 kHz. RESULTS: TEOEs of 574 ears were analyzed and the 2.0, 3.0 and 4.0 frequencies showed the highest pass percentages (94.1, 95.8 and 92.7% for protocols A, B and C respectively). Protocol C showed the highest pass percentage, achieving 96,9%. Nevertheless, the results obtained in Protocol B, which had the presence of three bands, showed a similar percentage to Protocol C (96,2%). Therefore, there was a not statistically significant difference between Protocols B and C. 1.0 kHz frequency achieved a percentage of only 9.9 in the tested ears. CONCLUSIONS: one has to research Protocols A, B and C with gold standard in order to verify which one suits best in terms of sensibility and specificity, even though the three protocols studied showed similar results in pass criteria for two or three frequency bands.