The early selection and use of an appropriate amplification system may be the most important aspect in the aural! rehabilitation of children who are hearing impaired. The main goal guiding the selection of hearing aids for children is to maximize their residual hearing through amplification in order to facilitate speech and language development. Seewald, Ross, and Stelmachciwicz (1987) have developed a speech-spectrum based procedure for selecting hearing aid characteristics for children referred to as the Desired Sensation Level Procedure (DSL) approach. This is an objective method which has been developed specifically for use with young pre-verbal children. I The purpose of the present study was to determine how closely the previously fitted amplification systems of profoundly hearing-impaired children approximated the amplification targets that would be prescribed for their hearing losses by the DSL method. The data were used to examine the feasibility and appropriateness of the DSL method for prescribing amplification for children with profound hearing impairments. Twenty amplification systems worn by profoundly hearing-impaired children were electroacoustically evaluated to discover how closely they approximated the DSL fitting criteria. Hearing thresholds were obtained for each subject for the frequencies 250, 500, 1000, 2000, and 4000 Hz using puretone audiometry. The DSL computer program was used to generate prescribed 2 cc. coupler targets at each frequency. The subjects' amplification systems were electroacoustically analyzed to determine the measured 2 cc. coupler response. Means and standard deviations were reported for prescribed and measured gain values at each frequency. Two-tailed t-tests were computed to determine if a difference exists between prescribed and measured gain. The tests were considered significant at the .05 level. A significant difference between the means was found at 2000 and 4000 Hz. Correlation coefficients were calculated at each frequency to determine if a predictable relationship between prescribed and measured gain existed. Correlation coefficients showed weak relationships between the two groups of data. These results showed the difficulty of meeting prescribed amplification targets, particularly in the high frequencies for children with profound hearing impairments.