This paper presents the development of an atmospheric N2O isotopocule model based on a chemistry-coupled atmospheric general circulation model (ACTM). We also describe a simple method to optimize the model and present its use in estimating the isotopic signatures of surface sources at the hemispheric scale. Data obtained from ground-based observations, measurements of firn air, and balloon and aircraft flights were used to optimize the long-term trends, interhemispheric gradients, and photolytic fractionation, respectively, in the model. This optimization successfully reproduced realistic spatial and temporal variations of atmospheric N2O isotopocules throughout the atmosphere from the surface to the stratosphere. The very small gradients associated with vertical profiles through the troposphere and the latitudinal and vertical distributions within each hemisphere were also reasonably simulated. The results of the isotopic characterization of the global total sources were generally consistent with previous one-box model estimates, indicating that the observed atmospheric trend is the dominant factor controlling the source isotopic signature. However, hemispheric estimates were different from those generated by a previous two-box model study, mainly due to the model accounting for the interhemispheric transport and latitudinal and vertical distributions of tropospheric N2O isotopocules. Comparisons of time series of atmospheric N2O isotopocule ratios between our model and observational data from several laboratories revealed the need for a more systematic and elaborate intercalibration of the standard scales used in N2O isotopic measurements in order to capture a more complete and precise picture of the temporal and spatial variations in atmospheric N2O isotopocule ratios. This study highlights the possibility that inverse estimation of surface N2O fluxes, including the isotopic information as additional constraints, could be realized.