We evaluated the performance of independent microbiology laboratories in Tokyo over an 18-year period of participation in the external quality assessment (EQA) program, and we estimated the impact of the EQA program. The study design was a longitudinal retrospective analysis of performance, including isolation, identification, and antimicrobial susceptibility testing of bacteria from simulated patients' samples, in "open" surveys compared with "blind" surveys. Independent microbiology laboratories, licensed by the Tokyo Metropolitan Government, have been subject to mandatory evaluation by the EQA program since 1982. Survey reports, correspondence, annual guidance meetings, and inspections are used as quality improvement strategies. The performance for identification in "blind" surveys was significantly worse than that in "open" surveys (P < 0.001). Poorly performing laboratories had common features, including inadequate supervision by physicians and lack of familiarity with the impact of variations on the use of the standards. However, there were improvements in the performance of identification of some pathogens. The performance of susceptibility testing has not yet reached the relatively high level seen for identification. Some of the smaller laboratories have been gradually acquired by commercial chains operating outside Tokyo. The EQA program has established a role both in regard to laboratory improvement and as an educational tool. However, the program lags behind these of other developed countries in regard to the practical sciences. The main problems in regard to laboratory improvements are a shortage of human resources in clinical microbiology, lack of standardization of laboratory methods, and the pressures of financial constraints in the Japanese medical insurance system.