Attributing nitrogen export to specific land use within heterogeneous catchments remains a challenge due to the spatio-temporal variability in conditions influencing the mobilization and fate of nitrogen species. This study demonstrates that the measurement of dual stable isotopes of nitrate, taken along with routine tributary measurement of nitrogen in nitrate (NO3--N) and ammonium (NH4+-N), aids in apportioning sources of the overall nitrogen load during wet periods. An inverse modeling technique was developed to estimate the land use-specific export rates of NO3--N and NH4+-N from the Caboolture River Catchment in Queensland, Australia. Measurements of nitrogen in streamflow at 51 locations during six sampling campaigns (May 2012 to April 2013) were made along with catchment geospatial data that was used to disaggregate sub-catchments into six land use fractions. A hydrological model was applied to compute the runoff from each fraction and water routing through the stream network. This data was used within a nitrogen mixing model with inclusion delta N-15(NO3) and delta O-18(NO3). The land uses specific export rate was computed inversely as the posterior of a Bayesian interference applied to the model. During higher rainfall periods when export rates were highest, the main land use exporting nitrogen was wetland (110 g/ha/day NO3--N, 27 g/ha/day NH4+-N) resulted from mineralization and nitrification of organic N, followed by urban (16 g/ha/day NO3--N, 2.3 g/ha/day NH4+-N). The advantage of using the dual isotopes in conjunction with the nitrogen concentration data was demonstrated by reduced uncertainty in the computed rates during the higher rainfall periods, relative to calculations without delta N-15(NO3) and delta O-18(NO3).