The purpose of this study is to present the meso-scale distribution of productivity of agricultural land (P. A. L.) in Japan and its change during the period from 1960 to 1975. P. A. L. is defined as gross product (thousand yen) of agriculture per agricultural land. The gross product of processed agricultural products were, however, excluded from the figure. The data for gross product of agriculture are obtained from the statistics on the agricultural income published by the Ministry of Agriculture, Forestry and Fishery. The book includes the amounts of products such as rice, wheat, barley, other cereals, peas, beeps, potatoes, sweet potatoes, vegetables, fruits, industrial crops, flower, sapling, and livestock. Agricultural lands includes cropland, orchard, pasture, and grassland pasture. These data for the analysis in 1960, 1965, 1970 and 1975 are taken from the Agricultural Census of Japan. The study area covers entire Japan excluding Okinawa prefecture due to the unavailability of data. The whole area is divided into 305 unit districts (Fig. 1), general nature of which were fully examined by Birukawa et al. (1964). The first step of the analysis was the preparation of maps of P. A. L, in each year. Then, the change of cereal distributions of P. A. L. were investigated by comparing these series of maps one another. Subsequently, the polynominal trend surface analysis was applied for analyzing the general distribution pattern and the change from 1960 to 1975. The general pattern of P. A. L. for each polynominal degree in each year, was also analyzed. In addition, the general patterns of P. A. L, were classified into types, and the residuals were considered as local components. The maps of standardized P. A. L. shows high scores in the fringe area of the metropolitan areas facing the Pacific Ocean and areas facing the Inland Sea, while low scores are in the several areas in the remote and secluded districts of Hokkaido, and the areas facing the Japan Sea (Fig. 2). It is clear that the difference in scores between the districts facing the Pacific Ocean and those facing the Japan Sea increased from 1970 to 1975. The number of areas with very high score increased significantly from 1965 to 1970. As a result, the number of districts with very high score is twice as numerous in 1970 compared to 1960. The increasing rate of areas with high score has been gradually getting higher. The number of districts with medium score decreased drastically from 1965 to 1970, and this tendency continued untill 1975. On the other hand, districts with low score gradually increased in number. Thus in 1975, there are 33 additional districts with lowscore compared with 1960. Regional variation of standardized P. A. L, has decreased in the recent census (Fig. 3). The types of change of P. A. L. gradually increased in number during this period, and the patterns of the change have been diversified. Fig. 4 shows the fitness of the trend surface to P. A. L.. There is a tendency of gradual increase of fitness from the first to the fifth degree in each year of analysis. Except for 1960, the fitness decreases at the 6th degree, and sharply increases in the 7th degree. Judging from maps of each year, the fitness in 1965 generally showed the low percentage, while the percentage were high in 1970. Comparing the results of variance analysis (Table 1) with the change of fitness of P. A. L., the general distribution patterns are investigated in this paper, at the 8th degree for 1960, and at the 7th degree for the remaining 3 census years (Fig. 5). The inspection of this figure shows the regions of maximum on the trend surface extending from Osaka to eastern and southern Shikoku, and from Tokyo to Shizuoka prefecture, and the region centered around Akita.