Wave propagation in sea-ice-covered oceans is a key component of polar air–ice–ocean interactions and plays an important role in regulating energy transport and sea-ice evolution in both the Arctic and the Southern Ocean. Under the influence of sea ice, wave energy attenuates and wave propagation speed may vary. A series of previous observational and numerical simulation studies have focused on the quantitative analysis of these processes. Nevertheless, how surface currents modulate waves travelling through ice zones remains unclear. As open-ocean wave-current interactions substantially modify wave features, this work quantifies current-modulated wave attenuation within polar marginal ice zones. In this study, the spectral wave model SWAN is used to hindcast wave fields over marginal ice zones in Polar regions, and to quantify the relative impacts of wind, sea-ice concentration, ice thickness, and ocean currents. Sensitivity numerical simulations are implemented to disentangle the contribution of each physical process. We benchmark the model results against field observations (e.g., PIPERS-2017) and reanalysis data (e.g., ERA5). Generally, the model overestimates wave heights under calm sea conditions and underestimates wave heights during storm events. However, the inclusion of ocean currents significantly improves the representation of large wave events. Spatially heterogeneous sea ice thickness in marginal ice zones is found to enhance wave attenuation under energetic conditions, while ocean currents act to modulate regional wave energy transport pathways. In terms of physical mechanisms, doppler shift and wave refraction induced by ocean currents during wave propagation can modify the spatial distribution of wave heights in ice-covered regions. These results demonstrate that apparent wave attenuation in sea-ice zones is governed by both wave-current and wave-ice processes rather than purely ice-driven damping. The findings provide new constraints on wave–ice–ocean parameterizations and highlight the need to revise traditional power-law-based attenuation frameworks in numerical wave models.