In recent years, there has been a growing interest to measure stimulus frequency otoacoustic emissions (SFOAEs) using sweep tones. While there are several advantages of the sweep-tone technique, one of the major problems with sweep-tone methodologies is the lack of an objective analysis procedure that considers and rejects individual noisy recordings or noisy segments. A new efficient data-driven method for rejecting noisy segments in SFOAE analysis is proposed and the normative features of SFOAEs are characterized in fifty normal-hearing young adults. The automated procedure involved phase detrending with a low-order polynomial and application of median and interquartile ranges for data outlier rejection from individual recordings. The SFOAE level and phase were analyzed using the least-squared fit method, and the noise floor was estimated using the error of the mean of the sweep level. Overall, the results of this study demonstrated the effectiveness of the automated noise rejection procedure and described the normative features of sweep-tone evoked SFOAEs in human adults.