Problems which must be addressed when dealing with sound measurement in an impulsive noise environment were discussed. First, there is a lack of any established dose/response relationship for such noise. Second, the dosimeters normally employed are not adaptable for use in an impulsive noise environment. Users of currently available equipment must assume that cutting the exposure time in half will reduce the auditory hazard by the same amount as if the duration remained the same and the exposure level were reduced by 3 or 5 decibels (dB). The author suggests for impulsive noise environments that the current equipment be modified to create a response appropriate to the required measurements. A modification is suggested which is based on an analysis of the response of simulated dosimeters to real impulsive noise samples. The major modification is that the new instrument should incorporate a much shorter time constant. The data presented suggest that a time constant of less than 30 milliseconds is needed for accurate estimations of exposure for impulsive noise components. No effect on continuous noise measurements will be noticed by reducing the time constant. Therefore any existing continuous noise data will be comparable with such data produced by the modified equipment measuring shorter time frames. While there is another solution involving the implementation of an averaging circuit using absolute value, this is more complex and offers no advantages over reducing the time constant.