Glen Canyon Dam has dramatically altered the physical environment (especially discharge regime, water temperatures, and sediment inputs) of the Colorado River. High-flow experiments (HFE) that mimic one aspect of the natural hydrograph (floods) were implemented in 1996, 2004, and 2008. The primary goal of these experiments was to increase the size and total area of sandbar habitats that provide both camping sites for recreational users and create backwaters (areas of stagnant flow in the lee of return-current eddies) that may be important as rearing habitat for native fish. Experimental flows might also positively or negatively alter the rainbow trout (Oncorhynchus mykiss) sport fishery in the clear tailwater reach below Glen Canyon Dam, Ariz., and native fish populations in downstream reaches (for example, endangered humpback chub, Gila cypha) through changes in available food resources. We examined the short-term response of benthic macroinvertebrates to the March 2008 HFE at three sites [river mile 0 (RM 0, 15.7 miles downriver from the dam), RM 62, and RM 225] along the Colorado River downstream from Glen Canyon Dam by sampling immediately before and then 1, 7, 14, and 30 days after the HFE. We selected these sites because of their importance to management; RM 0 has a valuable trout fishery, and RM 62 is the location of the largest population of the endangered humpback chub in the Grand Canyon. In addition to the short-term collection of samples, as part of parallel investigations, we collected 3 years of monthly (quarterly for RM 62) benthic macroinvertebrate samples that included 15 months of post-HFE data for all three sites, but processing of the samples is only complete for one site (RM 0). At RM 0, the HFE caused an immediate 1.75 g AFDM/m2 (expressed as grams ash-free dry mass, or AFDM) reduction of macroinvertebrate biomass that was driven by significant reductions in the biomass of the two dominant taxa in this reach-Potamopyrgus antipodarum (New Zealand mud snails) and Gammarus lacustris (scuds or side-swimmers)-and also biomass reductions of other common taxa (worms in the families Lumbricidae and Tubificidae). Invertebrate drift estimates during the HFE suggest that reductions in biomass of some taxa were because of export from the reach. Reductions in biomass of P. antipodarum and G. lacustris persisted at least 15 months after the HFE, when this study concluded, and coincided with a significant decline in the annual production of these taxa: P. antipodarum production of 11 to 13 g AFDM/m2/yr in two pre-HFE years versus 2 g AFDM/m2/yr in the post-HFE year, and G. lacustris production of 7 to 8 g AFDM/m2/yr in two pre-HFE years versus 3 g AFDM/m2/yr in the post-HFE year. There were not changes in invertebrate feeding habits in response to the HFE, as our 3-year dataset of invertebrate diets indicated no substantial changes. Our long-term analysis of the composition of the drift indicates that because of a reduction in P. antipodarum in the drift relative to digestible taxa, the quality of the drift as a food resource for fishes increased. At downstream sites, total assemblage biomass did not decline, likely because assemblages were dominated by blackflies (Simulium arcticum), which were not affected by the HFE. Similar to RM 0, G. lacustris and Tubificidae had significantly lower biomass after the HFE at RM 62 and RM 225. Chironomids were also significantly lower following the flood at both downstream sites. Our findings demonstrate that the effects of a HFE on invertebrates may persist up to at least 15 months in the clear tailwater below the dam, whereas in downstream reaches impacts were more short lived. If controlling the abundance of P. antipodarum is a goal of managers, our findings indicate that periodic HFEs on the order of every 2 to 3 years may be an effective strategy for meeting that goal. More frequent HFEs may cause a shift in the state of the benthic invertebrate assemblage of the tailwa