AbstractThis work developed three computational fluid dynamics (CFD) models of a full-scale oxidation ditch, one three-dimensional (3D) model and two one-dimensional (1D) models, to compare their predicted wastewater treatment performance. The models incorporated biokinetics through the Activated Sludge Model 1 (ASM-1) to predict the treatment performance of the ditch based on concentrations of pollutants: readily biodegradable substrate (SS), soluble ammonium ammonia nitrogen (SNH), and soluble nitrate nitrite nitrogen (SNO). When comparing the time series of the concentration of ASM-1 pollutants averaged over the ditch for 40 days, all three models displayed similar trends with slight differences in steady-state values, except for SNO. The steady-state value of SNO was greater by more than 150% for the 1D model than the 3D model. This difference is attributed to spatial heterogeneities in dissolved oxygen concentration predicted by the 3D model that were not captured by the 1D model, leading the latter to underpredict the denitrification process. Specifically, the spiraling flow around the aerators that plays an important role in determining the spatial distribution of dissolved oxygen cannot be represented in the 1D model.