The performance of nine commercially available personal noise dosimeter models was assessed through acoustical, electrical, and environmental evaluations. Twenty tests were conducted on each of three samples of the nine models evaluated. The tests included: rapid readout procedure, frequency response, acoustical calibrator accuracy, cutoff level, shifting levels, rise and decay time, exponential formula accuracy, fold over, integration accuracy, crest factor/square law capacity, battery life, battery indicator, dose storage, temperature sensitivity, magnetic field sensitivity, 115 decibel latch, sensitivity to atmospheric pressure variations, resistance to drop impact, cable integrity, and temperature endurance. Tests were conducted in accordance with a proposed NIOSH certification program for personal noise dosimeters. No single unit passed all the tests. Some units failed to function for the duration of the testing program. Except for units that never functioned properly, the percentage of tests passed by individual units ranged from 53.8 to 89.5. No attempt was made to assign a degree of importance to the individual tests. Depending on user requirements, certain test results were judged to be more critical than others. The authors conclude that manufacturers and users should look at each test on an individual basis. Specific requirements determine which tests carry the most impact.