A field research study was conducted to determine the actual noise attenuation and perceived comfort achieved by 40 noise-exposed industrial workers in 5 industrial workplaces wearing 4 different industrial hearing protection devices (HPDs) while on the job. Over 2 consecutive 3-week periods of HPD use, the study investigated the effects of 2 different HPD fitting procedures (subject-fit versus trained-fit) on the spectral field attenuation and user-rated comfort achieved with a user-molded foam earplug, a premolded, triple-flanged polymer earplug, a popular foam cushion earmuff, and an ear canal cap with compliant rubber earpods. Workers were pulled from their workplaces without prior knowledge of when they were to be tested and without re-adjusting the fit of their HPDs. Attenuation data were collected using psychophysical real-ear-attenuation-at-threshold measurement procedures as per the ANSI S12.6-1984 standard. Subjective comfort data were also obtained based on multi-dimensional bipolar rating scales. The results of statistical analyses indicated that when training for proper fitting was used, the earplugs significantly improved in noise protection (from 7.2 to 14.6 dB, depending on the frequency and the earplug) at frequencies of 125 - 1000, 6300, and 8000 Hz, whereas the earmuff and the ear canal cap were relatively resistant to the fitting effect. The training was most effective for the slow-recovery foam earplug over the 3-week period. For the comfort rating data, the foam earplug was again sensitive to the fitting effect, but the other HPDs were not. Among the 4 HPDs evaluated in the study, the canal cap protector was judged as the least comfortable HPD, while the other 3 HPDs yielded about the same perceived comfort. This research also showed that the overall field HPD protection afforded can be accurately predicted from single test band (i.e., centered at 500 or 1000 Hz) attenuation measurements. In addition, the field study demonstrated that laboratory simulation protocols designed to simulate field influences on HPD performance (used in the precursor laboratory study) may not be relied upon to yield accurate estimates of field performance of all HPDs. However, the estimates of field attenuation performance were more accurate for the earmuff than for the earplugs tested. Finally, this study demonstrated that the labeled manufacturers' single-number noise reduction ratings (NRRs) and frequency-specific data substantially overestimate the actual HPD attenuation performance achieved in the field. Consequently, on the basis of the results of this study, it appears that an appropriate, device-specific derating scheme to correct unrealistic labeled attenuation data is needed. / Ph. D.