Results of judgment tests on the noisiness of aircraft sounds published during the past decade [and cited by K. D. Kryter, J. Acoust. Soc. Amer.43, 344–361 (1968)] have been analyzed collectively to find the relative precision with which the judgments are explained by calculated perceived noise level, calculated loudness level, and Sound LevelsA, B, C,andD(formerlyN). For consistency, all levels have been recalculated, in accordance with current rules, from published noise spectra. Application of standard statistical procedures to the available data indicates that differences among the first three levels and the sixth, in their correlation with judged noisiness, are not statistically significant; only theBandClevels are clearly inferior as predictors of noisiness. A new weighting derived simply from two cascaded RC circuits, with corner frequencies at 2 and 8 kHz, leads to levels of better precision than any of the others, but adoption of this new weighting is not proposed without further study. Since Sound LevelApredicted the judged noisiness of the aircraft sounds as precisely as did calculated loudness level or perceived noise level, this readily measured sound level may well be applied to the rating of aircraft noise along with other noises in the community.