Diel dissolved oxygen (DO) data were used to characterize seasonal, inter-annual, and longitudinal variation in production and respiration for the James River Estuary. Two computational methods (Bayesian and bookkeeping) were applied to these data to determine whether inferences regarding DO metabolism are sensitive to methodology. Net metabolism was sensitive to methodology as Bayesian results indicated net heterotrophy (production < respiration) while bookkeeping results indicated net autotrophy (production > respiration). Differences in net metabolism among the methods was due to low seasonal variation in respiration using the Bayesian method, whereas bookkeeping results showed a strong correlation between production and respiration. Bayesian results suggest a dependence on allochthonous organic matter (OM) whereas bookkeeping results suggest that metabolism is dependent on autochthonous OM. This study highlights the importance in considering the method used to derive metabolic estimates as it can impact the assessment of trophic status and sources of OM supporting an estuary.