The assessment of biases in global streamflow reanalysis data is critical for practical applications, yet the influence of meteorological biases on streamflow biases remains unclear across large-sample catchments. Using observed hydrological data from 671 CAMELS catchments, this study examines and attributes biases in GloFAS global streamflow reanalysis. A panel regression approach is employed to quantify the empirical relationship between streamflow reanalysis biases and meteorological biases at seasonal scales, with bootstrap resampling used to estimate parameter uncertainties. Key findings include:(1) Low-flow biases are less sensitive to meteorological biases than high-flow biases, with meteorological influences on streamflow biases increasing nonlinearly at higher streamflow quantiles. (2) The influence of precipitation biases intensifies with increasing catchment area, soil depth, and precipitation seasonality, while weakening with lower snow-to-rain ratios and gentler slopes. (3) Potential evapotranspiration (PET) biases exhibit seasonal variations: in spring, they affect snowmelt-driven high flows, while in summer, they influence intermediate flows from snowmelt and baseflow in thin-soil catchments. (4) Catchments with high forest cover, vegetation-mediated evapotranspiration partially compensates for intermediate-flow biases in summer, demonstrating a hydrological buffering effect. This study reveals the nonlinear response of streamflow reanalysis biases to meteorological biases and elucidates the regulatory roles of catchment attributes and seasonal dynamics in bias propagation.