The objective of this study is to reveal the cause of noise annoyance based on its perceptual properties.The available physical factors extracted from the autocorrelation function (ACF) were used to explain the primary sensations and representative perceptual properties of road-vehicle noise.Two novel factors,EMD and ESM.used especially for measuring fluctuation and smoothness of the ACF envelope,were proposed.It should be pointed out that ifφ_1 was not the maximum value exceptφ(0),bothφ_(max) andτ_(max) were suggested to be added.Three primary sensations -loudness,pitch,and timbre have been acquired by applying the principal component analysis to ACF factors extracted from 232 noise samples.From 10 subjects, annoyance scale values were investigated using the magnitude estimation approach under wellcontrolled experimental conditions.It is realized that there certainly exist elementary factors for every primary sensation.Different ACF characteristics can induce diverse dominant sensation, which would control or determine the perceptual properties of the noise.The variability of dominant sensation and the priority difference in primary sensations will bring out subjective annoyance.Moreover,the mixture of elementary factors would confuse the auditory features, and induce subjective annoyance.In other words,it is the chaos of perceptual properties that cause and affect noise annoyance. The objective of this study is to reveal the cause of noise anoyance based on its perceptual properties. The available physical factors extracted from the autocorrelation function (ACF) were used to explain the primary sensations and representative perceptual properties of road-vehicle noise. Two novel Factors, EMD and ESM.used especially for measuring fluctuation and smoothness of the ACF envelope, were proposed. It should be pointed out that ifφ_1 was not the maximum value exceptφ (0), bothφ_ (max) andτ_ (max) were suggested to be added.Three primary sensations -loudness, pitch, and timbre have been acquired by applying the principal component analysis to ACF factors extracted from 232 noise samples. From 10 subjects, annoyance scale values ​​were investigated using the magnitude estimation approach under well controlled experimental conditions. It is realized that there will certainly exist elementary factors for every primary sensation. Different ACF characteristics can induce diverse dominant sens ation, which would control or determine the perceptual properties of the noise. variability of dominant sensation and the priority difference in primary sensations will bring out subjective annoyance. Moreover, the mixture of elementary factors would confuse the auditory features, and induce subjective annoyance. In other words, it is the chaos of perceptual properties that cause and affect noise annoyance.