The practice of large phosphate additions to agricultural land has resulted in an increase depletion of limited mineable rock phosphate resources and phosphate accumulation in soils. Especially, Andisols have distinct features with large capacities to sorb phosphate on their colloidal surfaces. In Tokachi district, Hokkaido, where is one of the largest agricultural areas in Japan, more than half of the total agricultural lands are covered by volcanic ashes, resulting in the most soils classify as Andisols. Since the soils are extremely rich in humus and allophanic compounds, they can strongly adsorb phosphate, which has been applied to promote crop growth in amounts exceeding those absorbed by the crops. Continued applications of surplus phosphate in fertilizers and other soil amendments will therefore enhance phosphate accumulation in the Andisols. Extensive studies have been conducted to characterize the reaction and mechanisms by which P is sorbed in Andisols. However, few studies have been conducted on the actual amounts and fate of phosphate accumulated in Andisols. The objectives in this study are (1) to compare the total and extractable amounts of phosphate in soils of an uncultivated land and an agricultural land with relation to soil depth and morphological features, and (2) to investigate effects of long-term soil fertilization and management on the phosphate accumulation in connection with soil physico-chemical properties and colloidal constituents. Soil samples were collected from two Andisol profiles in Tokachi district, Hokkaido, Japan, one of which were from an upland field, and another from an adjacent uncultivated land. These Andisols are classified into Typic Hapludands. The samples of the uncultivated site were collected from 0 to 100 cm depth every 5 cm as bulk samples and core samples by using 100 cm³ stainless cylindrical cores. The samples of the agricultural site were collected from 0 to 80 cm depth with the same method for the uncultivated site. The collected bulk samples were air-dried and passed through a 2 mm sieve. Total phosphorous contents (as P₂O₅) of the soils were determined by digesting the air-dried samples with concentrated nitric acid and perchloric acid mixtures, followed by the quantification of phosphorous in the digested solution by using ICP-AES. Inorganic phosphate were fractionated by sequential extraction: 1) extraction with 2.5 % acetic acid and 1 mol L^-1 ammonium chloride, 2) extraction with 1 mol L^-1 ammonium fluoride, 3) extraction with 0.1 mol L^-1 sodium hydroxide, determined by colorimetric methods and designated as Ca-P, Al-P, and Fe-P (as P₂O₅), respectively. The core samples were used to bulk density. Surface soil samples (0-18 cm) were collected from eight long-term (25 years) experimental plots in the Tokachi Agricultural Experiment Station, Hokkaido, Japan. The methods to determine the total and extractable phosphorous were as same as above. The amounts of total phosphorous in the uncultivated soils ranged from 0.58 to 1.50 g kg^-1. They were largest in the surface soil (0-5 cm) and larger in the upper-middle soils (25-40 cm) where phosphate absorption coefficients were relatively higher. Meanwhile, the total P in the upland soils ranged from 0.67 to 3.62 g kg^-1. They were rich in the Ap horizons (0-40 cm) and mostly exceeded 3 g kg^-1, more than half of which existed as Al-P. The positive correlation between total P and Al-P in the agricultural soils were observed, suggesting that most of surplus phosphate would retain by specific adsorption on active aluminum. Since the bulk density was extremely higher in the agricultural soils due to heavy machinery soil compaction, the differences in the total amounts of phosphorous based on the soil volume were more distinctive between two profiles, ranging from 0.43 to 0.93 kg m^-3 and from 0.57 to 3.16 kg m^-3 in the uncultivated and agricultural sites, respectively. In the lower part of the profiles (more than 50 cm depth), less difference in the total P between two profiles were observed. Our results showed that the phosphate originated from excessively applied fertilizers concentrated and remained in the upper part of the Andisol profile (within 50 cm depth) of the upland field. The total phosphorous in the surface soil collected from the plot, where only chemical fertilizers had been applied for 25 years, was 3.26 g kg^-1, and comparatively larger than that from no fertilizer plot. Long-term application of farmyard manure and incorporation of crop residues could increase in the total phosphorous. Especially, manure application contributed more to the increase in total P. The total P in the plot, where 30 Mg ha^-1 y^-1 of manure and fertilizers had been continuously applied, was 4.57 g kg^-1 and largest among the plots. The Al-P dominated in the inorganic phosphate fractions of all soil samples, followed by the Fe-P, and the amounts of the Ca-P was lowest. Long-term application of organic matter could increase in the total carbon contents and decrease in the allophane contents in the soils. However, the negative correlation was observed between the amounts of Al-P and the allophane contents, suggesting that the Andisols would have huge capacities to retain surplus phosphate. In conclusion, allophonic Andisols have incredible abilities to retain and accumulate phosphate derived from soil fertilization and organic management. Practical methods to increase an efficiency of phosphate and to reduce the amount of phosphate fertilizers applied to Andisols, and useful ways to release the phosphate accumulated in the Andisols under intensive agricultural production, should be investigated and developed for sustainable agriculture.