Operation of Sherburne Dam in northcentral Montana has typically reduced winter streamflow in Swiftcurrent Creek downstream of the dam and resulted in passage limitations for bull trout (Salvelinus confluentus). We defined an empirical relation between discharge in Swiftcurrent Creek between Sherburne Dam and the downstream confluence with Boulder Creek and fish passage geometry by considering how the cross-sectional area of water changed as a function of discharge at a set of cross sections likely to limit fish passage. With a minimum passage window of 15 x 45 cm, passage at the cross sections increased strongly with discharge over the range of 1.2 to 24 cfs. Most cross sections did not satisfy the minimum criteria at 1.2 cfs, 25 percent had no passage at 12.7 cfs, whereas at 24 cfs all but one of 26 cross sections had some passage and 90 percent had more than 3 m of width satisfying the minimum criteria. Sensitivity analysis suggests that the overall results are not highly dependent on exact dimensions of the minimum passage window. Combining these results with estimates of natural streamflow in the study reach further suggests that natural streamflow provided adequate passage at some times in most months and locations in the study reach, although not for all individual days and locations. Limitations of our analysis include assumptions about minimum passage geometry, measurement error, limitations of the cross-sectional model we used to characterize passage, the relation of Sherburne Dam releases to streamflow in the downstream study reach in the presence of ephemeral accretions, and the relation of passage geometry as we have measured it to fish responses of movement, stranding, and mortality, especially in the presence of ice cover.