Based on the recommendations of a 1992 conference on tuberculosis, the Centers for Disease Control and Prevention (CDC) established programs for upgrading mycobacteriology laboratories by providing them with monies and focused training. In 1991, state public health laboratories were surveyed to determine the methods they were using for primary Mycobacterium tuberculosis testing and their turnaround times for reporting testing results. A similar survey of nonstate laboratories participating in the National Laboratory Training Network-sponsored, M. tuberculosis-focused training programs was conducted from May 1992 to June 1993. In 1994, follow-up surveys of both the state- and nonstate-laboratory cohorts were conducted with the questionnaire from the initial survey plus additional questions that asked about interventions and changes occurring in the laboratory since the original survey. Although both cohorts showed increases in the percentages of laboratories meeting the recommended turnaround times for reporting M. tuberculosis testing results and using the recommended rapid methods for testing, generally, the increases made by the state laboratories were greater. By June 1994, all state laboratories were using a rapid method for M. tuberculosis isolate identification compared with 88% of the nonstate laboratories. The percentage of laboratories identifying isolates within the recommended 21 days also increased more in the group of state laboratories than in the group of nonstate laboratories (state laboratories, 22 to 73%; nonstate laboratories, 55 to 59%). Responses from the follow-up survey showed large differences in the percentages of laboratories that received CDC funding (state laboratories, 100%; nonstate laboratories, 6%) and participated in M. tuberculosis training (state laboratories, 98%; nonstate laboratories, 45%). These results indicate that adequate funding and focused training are critical in maintaining state-of-the-art mycobacteriology laboratories.