Land-applied manure is a significant source of N2O, but current estimates of emission are based on laboratory studies or field measurements from short-term application studies. Manure is often repeatedly applied to the same field for long periods, however. Our objectives were to determine annual N2O emission from a Typic Haploboroll soil that received 21 annual applications of solid feedlot manure at rates of 0, 60, 120, and 180 Mg ha(-1) (wet weight), and to relate the emission rates to environmental conditions. Nitrous oxide fluxes from soil were measured weekly from 10 Nov. 1993 until 27 Oct. 1994. Annual N2O emission increased with manuring rate and reached 56 kg N ha(-1) yr(-1) at the highest manuring rate (<1 kg N ha(-1) in the control). The N2O emission rate was not related to any single environmental factor, apparently because the N2O emission rate is not only controlled by rates of N2O production, but also by its rate of diffusion. The N2O emission rates were highest in early spring, but fluxes of N2O were also significant in winter, perhaps associated with freeze-thaw events. The N2O emission rates observed in our study are similar to N-fertilizer or combining N-fertilizer and manure application, but they are higher than other short-term studies with similar manure. These results probably reflect the accumulation of NO3 and organic matter from repeated manuring and suggest that the N2O emission from long-term manured soils may be underestimated by quantifying fluxes from short-term manuring studies.