Regional conditions of agriculture have been investigated every five years by the Ministry of Agriculture and Forestry in Japan and the statistic data were published as agricultural censuses. In this study, the data obtained in 1980, 1985, 1990 and 1995 were used for finding regional agricultural characteristics of 58 towns (administrative districts) in Oita Prefecture. Total land areas of this prefecture are 6,338km^2 and about 76km^2 are used for agricultural production. Most of the remaining parts are composed of mountainous area. 1) As a result of statistic principal component analysis, the first and second principal components are considered to make sense. The first principal component has characteristics expressing the activity level of agriculture. From the distribution pattern of the calculated scores, 58 towns in Oita Prefecture are found to be classified into five distinct groups according to the agricultural activity. 2) The regional characteristics of active level of agriculture were proved to depend on inter relationships between the variables. Among the variables, the scale of cultivated lands, the number of workers and the amount of agricultural income in each farmhouse have related most to the first principal component score. It was supposed that the present agricultural conditions could be grasped by means of the analysis stated above. 3) The variations of active level and statistic data of agriculture for time series showed that the five regional groups in Oita Prefecture have changed with different style during these 15 years. It is effective to analyze the causative variables in the long time course for studying the regional activating technique. 4) From the analysis of correlation between the topographic conditions and agricultural statistic data, it was clarified that the highest land levels of agricultural land in each district were closely related to the factors of the number of farm households possessing large scale cultivated lands and the number of part-time farm households having more agricultural income than non-farm income. The high regression coefficient showed that the topographic conditions obviously affect the agricultural activities.