Depending on the soil environment, selenium (Se) can exist as several species differing greatly in bioavailability. Characterization of soil Se reserves is thus necessary in assessing the nutritional supply of this essential element. In low-Se areas, Se fertilization is an option for securing adequate Se nutrition. Fertilization is, however, challenged by the unknown fate of the residual Se. In this study, we aimed to clarify the Se status of selenate-fertilized field soils by fractionating soil Se into five pools: salt-soluble (KCl), adsorbed (KH 2 PO 4 /K 2 HPO 4 ), organically associated (NaOH), elemental (Na 2 SO 3 ) and recalcitrant Se (NaOCl). Changes induced in these fractions by repeated application of low selenate doses were examined by comparing samples collected from the same locations in 1992 and 2004. The distribution of Se among the five fractions was relatively similar in all soils. On average, 1% of the amount of Se acquired was salt-soluble, 17% adsorbed, 39% organically associated, 14% elemental and 29% recalcitrant organic Se or metal selenides. The added Se was distributed among several fractions, leading to a relatively small change in each fraction. However, in mineral soils (n = 5), a significant increase in soil Se concentration was found between 1992 and 2004 in adsorbed, organically associated and recalcitrant Se, as well as in the total amount of Se acquired as a sum of the fractions. Our findings provide the first field-scale experimental support for the theoretical assumption of accumulation of residual selenate in acidic mineral soil in insoluble form.