The aim of this study was to clarify the atmosphere-paddy exchange of ammonia (NH 3 ) at a single-crop rice paddy field in central Japan. Measurements of NH 3 air concentrations were carried out using passive samplers at two heights above the ground surface (4.0 and 1.5 m) on a weekly basis over a four-year period. The mean NH 3 air concentrations ± standard deviations at 4.0 m were 2.5 ± 0.8 and 3.0 ± 1.0 (20°C, 1013 hPa) in the cropping and fallow seasons, respectively. Sharp increases in the NH 3 concentrations coincided with the application of poultry manure and post-harvest fieldburning activities; those values exceeded the 95% range of the weekly mean concentrations. The NH 3 exchange fluxes were calculated using a gradient method in which the flux was expressed as the product of the difference in air concentrations multiplied by the diffusion velocity between the two heights. The mean NH 3 exchange fluxes ± standard deviations were 0.016 ± 0.022 and 0.019 ± 0.115 μg N m -2 s -1 of deposition in the cropping and fallow seasons, respectively; where the weekly mean flux was assumed to be zero if the difference in the air concentrations between the two heights was not significant (p < 0.05). In general, the paddy field acted as a net sink of atmospheric NH 3 , with an annual net deposition of 7-9 kg N ha -1 yr -1 . However, the application of poultry manure in February 2008 induced a strong NH 3 emission of approximately 10 kg N ha -1 during three weeks, which nearly counterbalanced the annual deposition for an entire year. Passive samplers are a convenient option for long-term monitoring, although the weekly mean exchange fluxes of NH 3 had a systematic error; a case study showed a 66% overestimation for the weekly mean exchange flux.