This paper discusses the distribution of income and wealth, and poverty in Japan from the late Meiji Era to the late 1930s, using microdata on assessed income obtained for the collection of the local household tax called Kosūwari in selected areas. The analysis demonstrates several interesting results. First, Gini coefficients based on household income in these selected areas show that income inequality rose as incomes went up. Second, the decomposition of the Gini coefficient among different income levels suggests that the income share in middle income groups dropped markedly when income inequality worsened sharply. Third, decomposing the Gini by income sources for rural areas showed that a decline in non-agricultural incomes affected rising rural income inequality during periods of economic recession, while income from agricultural activities and income by renting out farm land increased inequality during periods of economic recovery. Fourth, the income share of labor in the non-agricultural sector showed a tendency to decline since Japan’s labor market was characterized by the continued presence of Lewisian surplus labor. The share of land in the total cost of agricultural production fell from the 1910s, while the share of labor remained relatively stable. Fifth, wealth was distributed more unequally than income. The Gini coefficient for rice field ownership ceased to increase in the 1910s when the share of land in the total cost of production began to decline as mentioned above. Finally, the relative poverty rate increased as average income levels rose and income inequality decreased. Based on local income data, the relative poverty rate for the whole country is estimated to have been around 20% in the 1920s and 1930s.