We investigated the effects of recent moisture history on the relative production of N2O and N2 during denitrification in soil from cropped and successional ecosystems. The soils were pedogenically identical but had been managed differently for the past decade. Sieved soils were amended with nitrate, glucose, and water. Long-wet and short-wet incubations received 80 and 0%, respectively, of prescribed water 2 d before incubation and the rest just before incubation. The N2O and N2 production and N2O mole fraction (N2O/[N2O + N2]) were measured using acetylene inhibition. The N2 production and soil 15N enrichment were measured by 15N-gas evolution. The response of N2O mole fraction to moisture history differed by ecosystem. Mean N2O mole fraction in the successional system was about the same for long-wet and short-wet treatments (0.34 and 0.33, respectively). For the cropped system, however, the N2O mole fraction was 0.36 for the long-wet and 0.90 for the short-wet treatment. Thus, in the cropped system a much smaller proportion of end product was N2O if soil had been wet for 2 d. For N2 fluxes, the isotope method gave the same pattern (r = 0.92) but only about one-third the magnitude, suggesting that N2 derived from two distinct pools. Differences in response of N2O mole fraction for successional and cropped soils may be due to differences in microbial communities. Further knowledge of ecosystem differences with respect to N2O mole fraction and recent moisture history may improve modeled estimates of local and global N2O fluxes.