During five years since the last survey of drug-resistant tuberculosis in Japan, features of tuberculosis in Japan have been changed. A nationwide survey was conducted by the Tuberculosis Research Committee of Japan. A total of 38 hospitals in various districts of Japan participated in the cooperative study. Each collaborating laboratory sent all mycobacterial cultures isolated during 1 June to 30 November, 1992 to the reference laboratory of the Committee, where species of the isolates were identified and drug susceptibility of Mycobacterium tuberculosis isolates were reexamined. The reference laboratory received a total of 1,236 cultures. Among them, 290 cultures were excluded from further examination by various reasons, such as contamination (52 cultures), non-viability (53), growth of nontuberculous mycobacteria (182) and other reasons (3). Thus, drug susceptibility test results were available for 946 cultures, including 26 cultures from non-Japanese persons. In the local laboratories, two methods, the absolute concentration method using 1% Ogawa egg slant (standard method, 26 hospitals) and its modified method using a microwell plate (microtiter method, 12 hospitals), were used for drug susceptibility testing, and the standard method was used in the reference laboratory. The results in the local laboratories were compared with those in the reference laboratory. The overall coincidence rate between drug susceptibility results reported from the local laboratories and those from the reference laboratory was 92.5%. A high coincidence rate (94.3%) was seen when the standard method was used in both local and reference laboratories. On the other hand, the coincidence rates between the results with the microtiter method in the local laboratory and those with the standard method in the reference laboratory were lower (standard method vs microtiter method; P < 0.01). Out of 19 hospitals, when the isolates were tested by the standard method 17 (89.5%) showed high coincidence rates (over 85%). Three hospitals using the microtiter method showed the coincidence rate over 90%, while other three showed lower rate (less than 80%) with high overestimation rates (over 19%), indicating that there are variations among facilities in performing the microtiter test. A part of the results concerning the resistance patterns to five antituberculosis drugs will be reported elsewhere.