A recent study revealed that most patients were wearing National Health Service hearing aids that were capable of exceeding their uncomfortable loudness level (ULL) measured clinically (Munro et al., 1996). However, there is little evidence to show that these clinical measurements are a valid indicator of real-world auditory discomfort. The aim of this study was to investigate the relationship between ULL and real-world discomfort. The study involved 20 adult subjects, aged 41-92 years, who had been fitted monaurally with an NHS hearing aid. ULLs were measured using a probe-tube microphone situated close to the eardrum. Individual real ear to coupler differences were added to the SSPL90 in order to predict the maximum power output (MPO) of the hearing aid at the eardrum. Subjects completed a questionnaire designed to rate the loudness of different environmental sounds. The results show that the more the MPO value exceeded the ULL value, the more likely it was that the subjects reported loudness tolerance problems to environmental sounds of long duration. There was a statistically significant correlation between the ULL and discomfort ratings for sounds of longer duration, such as traffic and wind noise; but not for shorter-duration sounds, such as door banging. Subjects did not express real-world auditory discomfort when the MPO value matched the ULL value. These findings support the argument for setting hearing-aid MPO close to ULL.