We investigated selected socioeconomic variables for incidence of tuberculosis and its rate of changes that might be relevant for the design of appropriate prevention and control programs. Retrospective ecological analysis was done to examine the association between eight socioeconomic measures from the 1992 census and both the average rate and the rate of change of standardized annual notification rates for tuberculosis from 1988 to 1997 for each of the 23 wards in Tokyo. Multivariate analysis identified the proportion of households with livelihood aid (p < 0.001), number of public bath per 100,000 person (p < 0.001), population density (p = 0.012), and proportion of households below standard house space (p = 0.024) as variables positively associated with average tuberculosis notification rates. The rate of change was negatively correlated with three variables: proportion of owner occupied households (p = 0.001), number of tatami (living space) per capita (p = 0.021), proportion of households with livelihood aid (p = 0.038). The results showed there were certain clear socioeconomic risk factors in tuberculosis transmission, and that it might be possible to investigate the relation between the rate of change and each socioeconomic risk factor effectively with this method.