Demand for water is increasing in many parts of the world, but removal of significant volumes of water from streams for irrigation, industry, and municipal supply has the potential to affect aquatic biota. Protecting the ecological value of streams is of increasing interest to water managers and the public. However, few experimental studies address the impacts of flow reduction on stream habitat condition and biota. We reduced discharge by 80 to 95% in 3 small, stony-bottomed streams (<4 m width), using weirs and diversion channels/pipes and quantified diel patterns of density of naturally drifting macroinvertebrates before and during a 2-mo period of continuous reduced flow. We also measured artificially induced passive invertebrate drift distance in control and impact reaches to determine the impact of flow reduction on drift distance. We measured the head capsule width of common drifting taxa to determine if larval size interacted with discharge to influence drift distance. Drift density of several taxa increased dramatically in the days following flow reduction, but drift returned to prereduction levels for the remainder of the reduced-flow period. Drift density of some taxa was elevated throughout the reduced-flow period. Few individuals drifted >1 or 2 m under reduced-flow conditions. Larval size influenced drift distance only for the mayfly Coloburiscus humeralis. We concluded that some taxa respond to the stress of reduced flow by drifting. Whether animals actively entered the drift or increased drift density was caused by increased benthic activity (i.e., invertebrates moving more) that, in turn, initiated increased passive drift entry, is unclear. Reduced discharge limits the distance an individual can travel in the water column. In small, shallow streams, drifting is unlikely to be a viable strategy for rapid escape of unfavorable conditions as a result of reduced discharge through flow abstractions.