The databases and models for the prediction of noise-induced permanent threshold shift (NIPTS) from industrial noise exposures are reviewed. Models available in 1973, compared with later models and data, are shown still to be reasonable. The effect of hearing conservation procedures on the acquisition of new data is discussed. Because of the impact of hearing conservation, new research focus is recommended in three areas: sex differences, newly hired individuals exposed for the first time in occupational noise, and the contribution of nonoccupational noise through the use of questionnaires and dosimetry.