Abstract 
The widths of western boundary currents (WBCs) have been investigated since the pioneering studies of Stommel (1948) and Munk (1950) on western intensification. Nevertheless, the governing dynamics remain opaque. Satellite observations have demonstrated that the widths of WBCs exhibit complex spatial variability. The magnitudes of the classical Stommel, Munk and inertial boundary layer scales are far smaller than the observed widths of WBCs. The Rhines scale has been found to be consistent with the observed width of WBCs in terms of both magnitude and along-stream variation, showing the influences of the eddies. However, these correlations are subject to modulation by topographic slope. The correlation between topographic slope and current width is generally negative. In circumstances where topographic slope varies continuously and substantially along the current path, the width response lags the topographic variation by approximately 2–3°. These findings have been corroborated by idealized numerical experiments in the MITgcm model. In summary, the Rhines scale provides a reasonable estimate of the fundamental width of WBCs. Furthermore, the topography modulates their width through a conditional, lagged downstream response. The details dynamic should be investigated further.