An investigation was conducted to determine the feasibility of implementing computerized audiometry in various clinical groups, using the Battery of Basic Computerized Audiometry Tests (BOBCAT). Reliability, validity and speed of execution were assessed as a function of hearing loss in a group of noise-exposed workers. Children and the aged were also included to represent potentially 'more-difficult-to-test' patients due to fluctuating attention, motivation and/or response behavior. Children were aged 7.5-12 years; seniors, 65-80 years. Reliability was assessed by calculating reliability coefficients between air conduction pure-tone thresholds (0.5, 1, 2, 3, 4 and 6 kHz) obtained under two test modalities namely, computerized audiometry and conventional testing performed by a small panel of trained examiners. Both procedures followed ANSI S3.21 1978 standards. Content validity was measured using measures of central tendency and correlations. Coefficients of reliability remained equally high across frequencies regardless of degree of hearing loss and group. As well, group means and correlations between conventional and computerized audiometry indicated that the two methods measured pure-tone hearing sensitivity with the same degree of accuracy; that is, within +/- 0.5 dB. Finally, speed of execution was found to be slower using BOBCAT, in particular with noise-exposed workers. This finding was interpreted as evidence that trained examiners have used shortcuts when they expect consistent and/or unbiased response behavior.