Direct loudness scaling has been known as an audiological tool for about four decades. Although numerous publications have shown its clinical importance, loudness scaling has not been used in audiology and hearing aid fitting until now. This might be due to the lack of audiometers equipped with loudness scaling devices as well as missing evidence for its clinical applicability. In this study normal data for a single-step direct scaling procedure were established and loudness determinations of 105 patients with sensorineural hearing losses collected. The results show that normal level loudness functions exhibit very similar shapes for narrow band stimulation in the frequency range of 500 to 4000 Hz. However, loudness scaling is affected by the gender of the subjects: females scale systematically louder than do males. In hearing-impaired subjects the slopes of the level loudness functions tend to decrease with increasing hearing loss, indicating negative recruitment. This holds particularly true in the high-frequency range, e.g. at 4000 Hz. Our long-term experience with single-step direct loudness scaling has proven its clinical feasibility in typical patients of an audiology unit. Loudness scaling will prove useful for the localization of hearing impairments, as it can be employed as a quantitative indicator of recruitment without any restrictive preconditions. In addition, it allows the evaluation of hearing aids and cochlear implants by frequency-specific gain measurement.