The occurrence and significance of the anammox (anaerobic ammonium oxidation) process relative to denitrification was studied in photosynthetically active sediment from 2 shallow-water estuaries: Randers Fjord and Norsminde Fjord, Denmark. Anammox accounted for 5 to 24% of N2 production in Randers Fjord sediment, whereas no indication was seen of the process in sediment from Norsminde Fjord. It is suggested that the presence of anammox in Randers Fjord and its absence from Norsminde Fjord is associated with differences in the availability of NO3- + NO2- (NOx-) in the suboxic zone of the sediment. In Randers Fjord, NOx- is present in the water column throughout the year and NOx- porewater profiles showed that NOx- penetrates into the suboxic zone of the sediment. In Norsminde Fjord, NOx- is absent from the water column during the summer months and, via assimilation, benthic microalgae may prevent penetration of NOx- into the suboxic zone of the sediment. Volume-specific anammox rates in Randers Fjord were comparable with rates measured previously in Skagerrak sediment by other investigators, but denitrification rates were 10 to 15 times higher. Thus, anammox contributes less to N2 production in Randers Fjord than in Skagerrak sediment. We propose that the lower contribution of anammox in Randers Fjord is linked to the higher availability of easily accessible carbon, which supports a higher population of denitrifying bacteria. Amplification of DNA extracted from the sediment samples from Randers Fjord using planctomycete-specific primers yielded 16S rRNA gene sequences closely related to candidatus Scalindua sorokinii found in the Black Sea by other investigators. The present study thus confirms the link between the presence of bacteria affiliated with candidatus S. sorokinii and the anammox reaction in marine environments. Anammox rates in sediment with intact chemical gradients were estimated using both 15N and microsensor techniques. Anammox rates estimated with microsensors were less than 22% of the rates measured with isotopes. It is suggested that this discrepancy was due to the presence of fauna, because the applied 15N technique captures total N2 production while the microsensor technique only captures diffusion-controlled N2 production at the sediment surface. This hypothesis was verified by consistent agreement between the methods when applied to defaunated sediments.