Global hydrological models (GHMs) have become an increasingly valuable tool in a range of global impact studies related to water resources. However, glacier parameterization is often overly simplistic or non-existent in GHMs. The representation of glacier dynamics and evolution, including related products such as glacier runoff, can be improved by relying on dedicated global glacier models (GGMs). In this study we test the hypothesis that coupling a GGM to a GHM can lead to increased GHM predictive skills and decreased GHM uncertainty through better glacier parameterization. To this end, the GGM GloGEM is coupled with the GHM PCR-GLOBWB 2 within the eWaterCycle II framework. For the years 2001-2012, the coupled model is evaluated against the uncoupled benchmark in 25 large (>50.000 km2) glacierized basins. Across all basins, the coupled model produces higher runoff throughout the melt season. In July and August, it ranges between 100.07% and 352% of the mean monthly benchmark runoff in lowly and highly glaciated basins respectively. The difference can primarily be explained by the inability of PCR-GLOBWB 2 to simulate snow redistribution and glacier retreat, causing an underestimation of glacier runoff. The coupled model better reproduces basin runoff observations primarily in highly glaciated basins, i.e. where the coupling has the most impact. This study underlines the importance of glacier representation in GHMs and demonstrates the potential of coupling a GHM with a GGM for better glacier representation and runoff predictions in glaciated basins.