Eight healthy students were exposed to intermittent noise to evaluate the effects of traffic noise on man. The experimental noise exposure conditions were as follows: Exp. 1: Control. Exp. 2: Peak level, 80 dB(A); rise-decay time, 1 sec; duration, 3 sec; and duty cycle 8 sec (on fraction 25%). Exp. 3: Peak level, 75 dB (A); rise-decay time, 500 msec; duration, 4.5 sec; and duty cycle, 8 sec (on fraction 50%). In all experiments, the exposure time was 14 hr. Temporary threshold shift (TTS) growth was followed for 14 hr while saliva cortisol levels were measured for a 24-hr period. Although the coefficient of the regression line of TTS growth recorded in Exp. 1 or Exp. 3 was not statistically significant, the value in Exp. 2 was statistically significant. The circadian rhythm pattern of the saliva cortisol level (during a 24-hr period, mean of 8 examinees) was observed in the three experiments. Essentially, the early morning level of saliva cortisol was high, while the daytime level was relatively low.