We tested an assumption of the instream flow incremental methodology that depth and velocity preferences are independent of streamflows. We had previously developed depth and velocity preferences (P[d] and P[v]) for juvenile (parr) steelheadOncorhynchus mykissat 0.86 m3/s in Morse Creek, Washington, and found parr distributed in microhabitats with higher combined depth–velocity preference (P[dv] = P[d] × P[v]) at a similar flow (0.69 m3/s). In the present study, we evaluated the relationship between fish distribution and combined depth–velocity preference using an independent data set from a higher flow (2.41 m3/s) in the adjacent stream segment. Most steelhead parr were distributed in microhabitats with high P[dv], consistent with distribution at 0.69 m3/s and significantly different than expected if fish distribution were independent of habitat preference (chi-square,P< 0.02). These results suggest that depth and velocity preferences are independent of flow.