Information regarding the relation of human temporary threshold shift (TTS) to properties of steady-state and intermittent noise published since the 1966 appearance of the CHABA damage risk contours is reviewed. The review focuses on results from four investigative areas relevant to potential revision of the CHABA contours including effects of long-duration exposure and asymptotic threshold shifts (ATS); equivalent quiet and/or safe noise levels; effects of intermittency; and use of noise-induced temporary threshold shift (NITTS) to predict susceptibility to noise-induced permanent threshold shift (NIPTS). These data indicate that two of three major postulates on which the original contours were based are not valid. First, recovery from TTS is not independent of the conditions that produced the TTS as was assumed. Second, the assumption that all exposures that produce equal TTS2 are equally hazardous is not substantiated. The third postulate was that NIPTS produced by 10 years of daily exposure is approximately equal to the TTS2 produced by the same noise after an 8-h exposure. Based upon several TTS experiments showing that TTS reaches an asymptote after about 8 h of exposure, the third CHABA postulate can be reworded to state the hypothesis that ATS produced by sound of fixed level and spectrum represents an upper bound on PTS produced by that sound regardless of the exposure duration or the number of times exposed. This hypothesis has a strong, logical foundation if ATS represents a true asymptote for TTS, not a temporary plateau, and if threshold shifts do not increase after the noise exposure ceases.