The leaching behavior of nitrogen was studied in single rice paddy production ecosystems in Tsukuba, Japan after 75 years of consistent fertilization regimes (no fertilizer, ammonium sulfate, a combination of composted rice straw with soybean cake, and fresh clover). During the 75-year period, management was unchanged with respect to rice planting density, irrigation, and net N fertilization for each field to which an N-source was added. Percolation water was collected, from May 2001 to April 2002, using porous suction cups installed in the fields at depths of 15, 40, and 60 cm. All water samples were taken to the laboratory for the measurement of both NH(4) ( + )-N and NO(3) ( - )-N concentrations using a continuous-flow nitrogen analyzer. The result indicated that there were significant differences in N leaching losses between treatments during the rice growing season. Total N leaching was significantly lower with the application of composted rice straw plus soybean cake (0.58 kg N ha( - 1)) than with ammonium sulfate (2.41 kg N ha( - 1)), which resulted in N leaching at a similar level to that with the fresh clover treatment (no significant difference). The majority of this N leaching was not due to NO(3) ( - )-N loss, but to that of NH(4) ( + )-N. The mean N leaching for all fertilizer treatments during the entire rice growing season was 1.58 kg N ha( - 1). Composted rice straw plus soybean cake produced leaching losses which were 65-75% lower than those with the application of fresh clover and ammonium sulfate. N accumulation resulting from nitrification in the fallow season could be a key source of nitrate-N leaching when fields become re-flooded before rice transplanting in the following year; particular attention should be paid to this phenomenon.