A new method for assessing the selected hearing aids was developed and applied 53 patients with hearing impairment. Our previous paper showed that the average confusion distance which was introduced in this method was useful for estimating the hearing aids. We carried out a follow-up survey in order to measure the usefulness of this method after 3 or 4 years practical hearing aid use. There were 34 patients out of 53 from whom we could obtain adequate information about their hearing aids. They were asked to rate their satisfaction with the recommended instrument in five degrees. Fifteen of the 34 patients who had purchased the recommended hearing aid rated it very satisfactory; four rated the aid satisfactory; eight rated the aid helpful; six rated the aid sometimes helpful; and one rated the aid unsatisfactory. The Spearman's rank correlation coefficient between the change in the speech discrimination score with and without a hearing aid and the user's evaluation was 0.34. The Spearman's rank correlation coefficient between the change in the average confusion distance and user's evaluation was 0.41. The value 0.34 is not significant, but the value 0.41 is significant at the 1% level of significance. Therefore, we think that the average confusion distance should be taken into consideration as well as the speech discrimination score at least for predicting successful hearing aid use. Finally we feel that there is need for the inclusive system which is able to estimate both the speech mediated communication ability and user's satisfaction.