Hypoxic conditions in aquatic systems and the occurrence of ‘dead zones’ increase worldwide due to man-made eutrophication and global warming. Nevertheless, for most environments existing monitoring activities are either inadequate or missing completely. Up to date hypoxia observation for a rapidly changing world has to (1) account for the appropriate temporal and spatial scales, (2) separate anthropogenic from natural drivers and long term trends from natural variations, (3) assess ecosystem response, (4) use modeling tools for generalization and prediction, and (5) share data and obtained knowledge. In 2009 HYPOX (www.hypox.net) started out as a pioneering attempt to improve hypoxia observation capacities addressing these requirements. HYPOX target ecosystems cover a broad range of settings (e.g., hydrography, biological activity, anthropogenic impact) and differ in their sensitivity towards change. Semi-enclosed basins with permanent anoxia (Black Sea, Baltic Sea), are included as well as seasonally or locally hypoxic land-locked systems (fjords, lagoons, lakes) and open ocean systems with high sensitivity to global warming (Arctic). HYPOX has built up oxygen observation capacities around Europe including long- term deployments of observatories as well as supplementary ship based observations. Modeling is used to synthesize findings and obtain an in-depth understanding of hypoxia causes and consequences. In order to integrate the collected information into a global oxygen observing system, results are disseminated through the HYPOX portal following GEOSS data sharing principles. The presented work will introduce the characteristics of the selected sites, give an overview of the scientific approach of HYPOX and highlight some results.