Forty-one stock organisms and 581 fresh clinical isolates were used in comparing the Minitek system to conventional tubed media to determine if this system is feasible and accurate for a high-volume clinical microbiology laboratory. In addition to comparison with tubed media, the following parameters were tested: (i) reproducibility of disks, (ii) the effect of variation in inoculum size, (iii) the effect of the age of culture, and (iv) the effect of predispensing disks. A total of 5,947 disks were compared with tube reactions; 95.9% agreed on the initial reading. After repeating questionable tube and disk reactions, this agreement rose to 98.2%. Using the disk reactions and the identification schema currently in use with tubed media, 94.9% of the Minitek final identifications agreed with those of tubed media; 1.5% gave differing identifications and 3.6% were inconclusive. This system is accurate, inexpensive, flexible, and convenient.