Dissolved organic matter (DOM) is ubiquitous in natural waters and plays a central role in the biogeochemistry of riverine, estuarine and marine environments. The heteroatomic fraction of DOM consists mainly of nitrogen, phosphorous, and sulfur. While the biogeochemical cycling of C, N, and P has been intensely studied, dissolved organic sulfur (DOS) has been only marginally addressed. Nevertheless, it is an essential element for marine primary production and organic sulfur compounds play a critical role in biogeochemistry, ecology, and climate processes. The analysis of DOS quantity and distribution in marine environments as well as its chemical characterization is of urgent need to further understand the underlying processes of DOS biogeochemistry. This study is based on more than 600 samples from different marine environments, from the surface to the deep ocean, and thus, represents the so far most comprehensive dataset of DOS in the ocean. Within this thesis, a basin-scale distribution of extractable DOS in the East Atlantic Ocean, the Atlantic sector of the Southern Ocean, and the Weddell Sea is presented. A first conservative global inventory of the marine non-volatile DOS stock was calculated using a combination of state-of-the-art approaches. The results suggest that the marine DOS inventory is by far the largest oceanic reservoir of organic sulfur (OS), exceeding the atmospheric stock and the volatile marine DOS by several orders of magnitude. Decades of research on the marine DOS cycle have focused on only ~2 % of the total OS inventory (DMSP cycle). The marine DOS distribution and stoichiometry was compared to radiocarbon age of dissolved organic carbon (DOC), suggesting a primarily autochthonous biogenic origin and an active involvement of this DOS in the microbial loop - similar to organic nitrogen. The contribution of the sulfur-containing amino acid methionine to extractable DOS was found to be only ~2 %. Additionally, ultrahigh-resolution mass spectrometry identified a diverse suite of sulfur containing molecular formulas and their relative contribution suggested shorter residence times relative to DOC supporting the active involvement of DOS in biogeochemical cycles, ecological processes and ecosystem functions. Rivers are also important sources of (sulfur containing) organic matter to marine environments. Here I quantified DOS and trace metals along a salinity gradient and explored the complexation of trace metals with organic matter, in particular with sulfur-containing compounds. Despite decreasing DOM concentrations along the salinity gradient due to estuarine mixing, we showed that DOS is degrading faster than DOC, underpinning the active involvement of DOS in biogeochemical cycles and ecological processes. We found indication for complexation of trace metals with (sulfur containing) organic ligands and confirmed that the stability of metal-organic complexes followed the Irving-Williams order. The results of this thesis challenge current views of DOS dynamics and suggest that there remain major gaps in our understanding of the marine sulfur cycle, which will be of interest to ocean biogeo-) chemists, atmospheric scientists, microbial ecologists, and ocean-/climate modelers.