In the southwestern United States, increases in climate variability are leading to dramatic fluctuations in stream discharge, including atypical large floods and long dry spells. This hydrological variability will affect aquatic biota, including aquatic invertebrates. National parks, as areas protected from most anthropogenic impacts, can be useful to examine the effects of hydroclimate variability on macroinvertebrates while minimizing confounding factors. We studied changes in invertebrate community structure in five national parks in southwestern United States using a 10-year biomonitoring dataset spanning from 2012 to 2022. Macroinvertebrate data were collected from perennial streams in national parks in the Sonoran Desert and Southern Plains ecoregions of Arizona and New Mexico. We found a total of 228 invertebrate taxa for this time period, including 122 taxa at Gila Cliff Dwellings National Monument (GICL), 145 taxa at Montezuma Castle National Monument, 66 taxa at Pecos National Historic Site (PECO), 68 taxa at Tumacacori National Historic Site, and 124 taxa at Tuzigoot National Monument. Invertebrate communities were compositionally distinct among parks, with biogeographic separation between communities in Arizona and New Mexico. Across the 10-year period, the interannual coefficient of variation in taxonomic richness within parks was low, averaging 23% and ranging from 19% (PECO) to 28% (GICL). Changes in taxonomic richness from year-to-year were not synchronized across parks, indicating a relative lack of region-wide predictable changes through time. The only exception was for the 2017-2018 water year, where richness increased at four of the five parks. This relative lack of synchronized changes across the region suggests that local hydrological and habitat conditions are likely driving the large amount of interannual variation in richness that we observed. Future analyses, including stream flow variability and statistical modeling of species abundances, will provide additional insight into these patterns.