This Bureau of Mines report evaluates audio dosimeters, 3 sample units from each of 10 currently marketed brands. Each dosimeter was treated as a "black box" with the acoustic stimulus as the input and the accumulated dosage as the output. The dosimeter microphones were exposed to a sound field generated in an anechoic chamber. Characteristics evaluated include frequency response (100 to 8,000 hz), level-time tradeoff rate (88 to 117 db), status of calibration, crest-handling capability, and power consumption rate. Experimental errors revealed by replication of the tests were found to be generally much smaller than variations in performance among the samples of any brand, and possibilities of further reducing the magnitude of these errors are discussed.