We quantified the removal of fixed nitrogen as N 2 production by anammox and N 2 and N 2 O production by denitrification over a distance of 1900 km along the coasts of Chile and Peru, using short-term incubations with 15 N-labeled substrates. The eastern South Pacific contains an oxygen minimum zone (OMZ) characterized by an anoxic, nitrate- and nitrite-rich layer of , 200-m thickness below 30–90 m of oxic water. Anammox and denitrification were almost exclusively recorded when the in situ O 2 concentration was below detection, indicating that the induction of these processes is highly oxygen sensitive. Anammox was detected in 70% of the samples from anoxic depths. Denitrification was detected in fewer samples, but maximum rates were an order of magnitude higher than those of anammox. In our incubations denitrification was responsible for 72% of the total N 2 production and 77% of the total removal of fixed nitrogen including N 2 O production. However, at the individual depths it could be one or the other process that was responsible for all of the nitrogen removal. Anammox activity was highest just below the oxic–anoxic interface and declined exponentially with depth, whereas no depth dependence was discerned for denitrification. Denitrification resulted in net production of N 2 O in some of the samples and consumption of added 15 N 2 O in others. Together with the accumulation of NO 2 - this indicates that denitrification must be seen as a sequence of individually regulated reactions, each of which may start and stop depending on the electron donor input, while anammox is much less variable. The highly patchy distribution of denitrification contributes to explain the apparent imbalances between ammonium sources and sinks suggested by previous 15 N-based studies in OMZs