The instream flow incremental methodology predicts the potential amount of habitat in a stream, rather than fish biomass or numbers. We introduced rainbow trout into stream channels next to the Waitaki River, South Island, New Zealand, and subsequently maintained a constant flow in these streams. The biomass of rainbow trout in individual riffles and pools of the streams was determined and related to the amount of usable area in these habitats calculated using the incremental methodology. Regardless of fish stocking density, rainbow trout biomass in late summer and early winter was not correlated with the amount of usable habitat. Users of the incremental methodology should realise that the method cannot be expected to predict fish biomass or numbers when fish are not limited by the amount of space available to them. When fish are food‐limited, as apparently occurred in our streams, relationships between usable area, fish food organisms, and fish biomass must be understood before one can predict flow change impacts.