A range of proxy approaches have been used to reconstruct short-term changes to Earth’s hydrological cycle during the early Eocene hyperthermals. However, little is known about the response of Earth’s hydrological and biogeochemical systems to long-term Cenozoic cooling, which began following the Early Eocene Climatic Optimum (53.3 – 49.4 million years ago; Ma). Here, we use the molecular distribution and isotopic composition of terrestrial biomarkers preserved in marine sediments of ODP Site 913, East Greenland, to develop a long-term record of high-latitude hydrological change between 50 and 34 Ma. There is a marked decline in the concentration of conifer-derived diterpenoids and angiosperm-derived triterpenoids during the Eocene. As the input of wind-blown conifer pollen remains stable during this interval, this implies that decreasing di- and triterpenoid concentrations reflect declining influence of fluvial inputs – and perhaps terrestrial runoff – throughout the Eocene. Branched GDGTs and bacterial-derived hopanes indicate an increased input of soil- and kerogen-derived organic matter, respectively, after 38 Ma. This coincides with evidence for ice rafted debris and suggests input of organic matter via glacial processes. This also implies some continental glaciation occurred on East Greenland in the middle-to-late Eocene. Leaf wax hydrogen isotopes extending throughout this section – the first such long-term record from the Paleogene - indicate that precipitation δ2H was persistently higher than that of modern coastal Greenland, consistent with warmer ocean source waters and enhanced poleward moisture transport. Non-intuitively, however, this effect appears to have been smallest during the warmest part of the record, and higher δ2H values occur in the middle Eocene. Although interpretation of these hydrogen isotope trends is unclear, they clearly indicate – alongside the changes in biomarker abundances – a perturbed hydrological cycle through the Eocene in coastal Greenland. More long-term records are required to ascertain if this represents regional or global hydrological reorganisation.