In this study, we measured nitrous oxide (N2O) emissions from plots of fall-planted hairy vetch (HV, Vicia villosa) grown as a green nitrogen (N) source for following summer forage crabgrass (Digitaria sanguinalis). Two treatments were compared: (i) HV grown solely as green manure where all biomass was incorporated by tillage, and (ii) harvesting of aboveground HV biomass prior to planting of crabgrass. Fluxes of N2O were measured with closed chamber systems on 27 dates during a 2-month growth period of crabgrass after the termination of HV in early May. At termination, the average aboveground biomass yield of HV was 4.6 Mg ha−1 with 146 kg N ha−1 content. The N2O emissions were as high as 66 g N2O-N ha−1 day−1 on day 1 after HV incorporation, but reached close to zero within a week. Emissions of N2O increased with subsequent rainfall and irrigation events from both treatments but emission peaks were not observed during the rapid growth of crabgrass. Two-month cumulative emission of N2O (mean ± s.e., n = 4) from HV incorporated plots (921 ± 120 g N2O-N ha−1) was three times (P < 0.05) of HV harvested plots (326 ± 30 g N2O-N ha−1). However, crabgrass biomass yields, N concentrations and total biomass N uptake were decreased significantly by harvesting HV. In conclusion, the results suggested that whereas removal of HV biomass for use as forage may significantly reduce N2O emissions, quantity and quality of the following recipient crops may be constrained.