Noise levels are truly continuous in relatively few occupations, with some degree of intermittency the most common condition. The sound levels of intermittent noise are often referred to as non-Gaussian in that they are not normally distributed in the time domain. In some conditions, intermittent noise affects the ear differently from continuous noise, and it is this assumption that underlies the selection of the 5-dB exchange rate (ER). The scientific and professional communities have debated this assumption over recent decades. This monograph explores the effect of non-Gaussian noise on the auditory system. It begins by summarizing an earlier report by the same author concentrating on the subject of the ER. The conclusions of the earlier report supported the more conservative 3-dB ER with possible adjustments to the permissible exposure limit for certain working conditions. The current document has expanded on the earlier report in light of the relevant research accomplished in the intervening decades. Although some of the animal research has supported the mitigating effect of intermittency, a closer look at many of these studies reveals certain weaknesses, along with the fact that these noise exposures were not usually representative of the conditions under which people actually work. The more recent animal research on complex noise shows that intermittencies do not protect the cochlea and that many of the previous assumptions about the ameliorative effect of intermittencies are no longer valid, lending further support to the 3-dB ER. The neurologic effects of noise on hearing have gained increasing attention in recent years because of improvements in microscopy and immunostaining techniques. Animal experiments showing damage to auditory synapses from noise exposures previously considered harmless may signify the need for a more conservative approach to the assessment of noise-induced hearing loss and consequently the practice of hearing conservation programs.