Several hearing aid prescription procedures specify frequency gain functions in terms of the desired levels of amplified speech. The most direct method for verifying hearing aid fittings based on these procedures requires a measurement of amplified speech in the ear canal. This paper describes the evaluation of a probe microphone measurement procedure designed to measure amplified speech to verify hearing aid fittings using the MSUv3 prescription procedure. With minor modifications, the same protocol could be used with other prescriptive procedures. Hearing aids were fitted to a group of subjects with sensorineural hearing impairment. Data derived from the fittings were analyzed to assess several issues relating to the procedure itself and to prescriptive fittings in general. The main results were: (1) with over the ear hearing aids, most prescriptions can be matched with an RMS error of 5 dB or less through the frequency range from 500 to 2500 Hz; (2) even though actual fittings usually do not perfectly correspond with their prescriptions, differences among frequency gain prescriptions are preserved in hearing aid fittings if the RMS error of the fitting is 5 dB or less; (3) if sound field stimuli presented to the hearing aid are precisely controlled, an in situ output verification method produces valid results; (4) when hearing aids are fitted with gain similar to that prescribed by the MSUv3 procedure, the maximum comfortable speech input level is typically about 72 dB; (5) the SSPLSO prescription generated by the MSUv3 procedure overestimates desired SSPLSO by 7 dB on average.