The purpose of this study was to determine whether different release times, as implemented in a commercial two-channel AGC hearing aid, would result in differing speech intelligibility performance, user preference, or use time. In experiment one, 14 subjects were fitted with a two-channel multi-memory AGC hearing aid. Four memories were programmed to have identical frequency responses and output limitation characteristics. Only the release times were varied, with the low channel/high channel set as follows (in ms): 20/35, 20/150, 100/35, 500/7. Results obtained from the NST (+5 S/N), magnitude estimations of intelligibility, and data-logging of use time did not show any release-time pair to be superior to any other. In experiment two, 10 subjects participated in a forced-choice, paired-comparison procedure using the same release-time pairs from experiment one. Auditory stimuli consisted of three input levels, consisting of speech, speech in noise, and music. Results indicated no release-time pair to be superior in any listening condition. Results may be explained, in part, by the use of a curvilinear compression circuit and the milder hearing loss exhibited by the subjects. Future investigation of the effect of release-time variation should be carried out on circuits with adjustable compression parameters (and fixed compression ratios) with listeners exhibiting different degrees of hearing loss.