Effects of aquifer travel time on nitrogen reaction and loading to Popponesset Bay, a eutrophic coastal embayment on western Cape Cod, Massachusetts, are evaluated through hydrologic analysis of flow and transport. Approximately 10% of the total nitrogen load to the embayment is intercepted by fresh water ponds and delivered to the coast by connecting streams. For the nitrogen load not intercepted by ponds, we compare two steady-state methods of analyzing nitrogen loss in the aquifer, one using a constant-loss factor and the other time-dependent loss rates. The constant-loss method, which assumes that all similar land uses have the same per unit area loading rate to surface water regardless of location within the watershed, predicts that 42% of the nonpond watershed nitrogen load originated within the zero to 2 yr time-of-travel zone, which is 40% of the contributing area. The time-of-travel loss method calculates loss rates based on aquifer travel times and denitrification reaction kinetics, evaluated separately for carbon-unlimited and carbon-limited cases. Time-of-travel loss calculations for percent of nonpond load that originated within the area of < 2 yr aquifer residence time are 64% when carbon is not limiting, but only 49% when carbon limitation is included, not greatly different from the constant-loss method. A feature of the kinetics used is that carbon (and the denitrified nitrogen) is lost rather quickly in the aquifer travel path, after which carbon limitation stops denitrification altogether. Carbon limitation causes the time-of-travel loss model to approximate the constant-loss model such that in most of the watershed, a nearly constant fraction of the nitrogen input is lost in both models.