The ln(Zr/Rb) count ratio derived from X-ray fluorescence (XRF) core scanning holds potential as a high-resolution tracer for grain-size variations of glaciomarine sediments. To evaluate this approach, we conducted high-resolution grain-size measurements, together with Rb and Zr measurements by XRF core scanning and Inductively Coupled Plasma-Mass Spectrometry (ICP-MS), on a series of sediment cores from different regions of the Southern Ocean. We find that downcore changes of the ln(Zr/Rb) count ratio from XRF core scanning are consistent with Zr/Rb concentration ratios derived from ICP-MS analyses, even though Rb and Zr counts deviate significantly from concentrations due to specimen and matrix effects. The ln(Zr/Rb) count ratio displays discrepancies with the bulk mean grain-size, but correlates well with the mean grain-size of the sediment fractions that do not include unsorted coarse IRD (i.e. IRD-corrected mean grain-size). These observations are supported by evidence from a grain-size separation experiment, which indicates that Zr and Rb are concentrated in different grain-size fractions. Consistent with its lack of sensitivity to coarse grain-size fractions, the ln(Zr/Rb) ratio records similar trends to the sortable silt percent (SS%) and sortable silt mean (SSM) grain-size. Universal gradients exist in plots of ln(Zr/Rb) versus SS% (34.1), and ln(Zr/Rb) versus SSM (12.7), such that the ln(Zr/Rb) ratio provides a convenient way to estimate the magnitude of changes in SS% and SSM. Overall, our results support the use of the ln(Zr/Rb) ratio as an indicator of bottom current strength in cases where the sediment is current-sorted.