According to the operational definition adopted at Wolfheze Workshop held in March 1991 (Table 1), tuberculosis low incidence countries are defined as those where the incidence of all forms of tuberculosis is less than 1 per 100,000. The incidence was 41.9 in 1990 in Japan, so that Japan cannot be considered as a low incidence country. Why is the incidence of tuberculosis so high in Japan? What part of population is affected by the tuberculosis disease? To solve these questions and to make clear the mode of development of tuberculosis, a study was carried out. Fortunately, both the annual risk of tuberculosis infection and the prevalences of persons with healed and/or fibrotic lesion in the lung are known from the results of the National Tuberculosis Prevalence Surveys carried out in 1963 and 1973 as shown in Figures 1 and 2. From these data, the number of newly infected within 5 years and that of remote infections by age group were estimated. The former was divided into BCG vaccinated and non-vaccinated and the latter into persons with fibrotic lesion, with healed foci and without abnormality in the lung. (Table 2). The rates of development of tuberculosis disease according to the X-ray findings of the lung were observed at the five-years follow-up study of all the examinees at the Prevalence Survey carried out in 1968. The ratio of development of the disease in those with recent infections, remote infections with fibrotic lesion, those with healed foci and those without abnormality was estimated as 20 : 10 : 2 : 1. (Table 3)(ABSTRACT TRUNCATED AT 250 WORDS)