Notes: 1. The microbial metabolism of organic matter in rivers has received little study compared with that of small streams. Therefore, we investigated the rate and location of bacterial production in a sixth-order lowland river (Spree, Germany). To estimate the contribution of various habitats (sediments, epiphyton, and the pelagic zone) to total bacterial production, we quantified the contribution of these habitats to areal production by bacteria.2. Large areas of the river bottom were characterized by loose and shifting sands of relatively homogenous particle size distribution. Aquatic macrophytes grew on 40% of the river bottom. Leaf areas of 2.8 m2 m−2 river bottom were found in a 6.6 km river stretch.3. The epiphyton supported a bacterial production of 5–58 ng C cm−2 h−1. Bacterial production in the pelagic zone was 0.9–3.9 μg C L−1 h−1, and abundance was 4.0–7.8 × 109 cells L−1. Bacterial production in the uppermost 2 cm of sediments ranged from 1 to 8 μg C cm−3 h−1, and abundance from 0.84 to 6.7 × 109 cells cm−3. Bacteria were larger and more active in sediments than in the pelagic zone.4. In spite of relatively low macrophyte abundance, areal production by bacteria in the pelagic zone was only slightly higher than in the epiphyton. Bacterial biomass in the uppermost 2 cm of sediments exceeded pelagic biomass by factors of 6–22, and sedimentary bacterial production was 17–35 times higher than in the overlying water column.5. On a square meter basis, total bacterial production in the Spree was clearly higher than primary productivity. Thus, the lowland river Spree is a heterotrophic system with benthic processes dominating. Therefore, sedimentary and epiphytic bacterial productivity form important components of ecosystem carbon metabolism in rivers and shallow lakes.6. The sediments are focal sites of microbial degradation of organic carbon in a sand-bottomed lowland river. The presence of a lowland river section within a river continuum probably greatly changes the geochemical fluxes within the river network. This implies that current concepts of longitudinal biogeochemical relationships within river systems have to be revised.