Bacterioplankton-driven turnover of the algal osmolyte, dimethylsulphoniopropionate (DMSP), and its degradation product, dimethylsulphide (DMS) the major natural source of atmospheric sulphur, were studied during a Lagrangian SF 6 -tracer experiment in the North Sea (60 o N, 3 o E). The water mass sampled within the euphotic zone was characterised by a surface mixed layer (from 0m to 13-30m) and a subsurface layer (from 13-30m to 45-58m) separated by a 2 o C thermocline spanning 2m. The fluxes of dissolved DMSP (DMSPd) and DMS were determined using radioactive tracer techniques. Rates of the simultaneous incorporation of 1 4 C-leucine and 3 H-thymidine were measured to estimate bacterioplankton production. Flow cytometry was employed to discriminate subpopulations and to determine the numbers and biomass of bacterioplankton by staining for nucleic acids and proteins. Bacterioplankton subpopulations were separated by flow cytometric sorting and their composition determined using 16S ribosomal gene cloning/sequencing and fluorescence in situ hybridisation with designed group-specific oligonucleotide probes. A subpopulation, dominated by bacteria related to Roseobacter-(α-proteobacteria), constituted 26-33% of total bacterioplankton numbers and 45-48% of biomass in both surface and subsurface layers. The other abundant prokaryotes were a group within the SAR86 cluster of γ-proteobacteria and bacteria from the Cytophaga-Flavobacterium-cluster. Bacterial consumption of DMSPd was greater in the subsurface layer (41nMd - 1 ) than in the surface layer (20nMd - 1 ). Bacterioplankton tightly controlled the DMSPd pool, particularly in the subsurface layer, with a turnover time of 2h, whereas the turnover time of DMSPd in the surface layer was 10h. Consumed DMSP satisfied the majority of sulphur demands of bacterioplankton, even though bacterioplankton assimilated only about 2.5% and 6.0% of consumed DMSPd sulphur in the surface and subsurface layers, respectively. Bacterioplankton turnover of DMS was also faster in the subsurface layer (12h) compared to the surface layer (24h). However, absolute DMS consumption rates were higher in the surface layer, due to higher DMS concentrations in this layer. The majority of DMS was metabolised into dissolved non-volatile products, and bacteria could satisfy only 1-3% of their sulphur demands from DMS. Thus, structurally similar bacterioplankton communities exerted strong control over DMSPd and DMS concentrations both in the subsurface layer and surface mixed layer.