In aquatic plant-dominated stream and river reaches, ecosystem metabolism can be much higher than in other rivers, but few metabolism studies exist for medium and large rivers like these. In the lower Yakima River (Washington State, USA), dampened spring peak flows have led to the proliferation of water stargrass (Heteranthera dubia). This native aquatic plant dominates the ecosystem, often growing across the full width of the river. Large spring peak flows can remove water stargrass biomass, potentially reducing river metabolism. We tested the effect of spring flow by comparing metabolism during the growing season between a high-flow, low-water stargrass year and 5 years with typical low flows and high water stargrass biomass. We estimated metabolism with the 1-station Bayesian inverse modeling approach, using dissolved oxygen (DO) data from a U.S. Geological Survey monitoring location and from PME miniDOT DO sensors. In the low-flow years, mean summer gross primary production (GPP) was 36.5 g O₂ m⁻² d⁻¹ (range: 23.7 – 41.3 g O₂ m⁻² d⁻¹) and mean summer ecosystem respiration (ER) was -25.4 g O₂ m⁻² d⁻¹ (range: -15.9 – -28.4 g O₂ m⁻² d⁻¹). The high flow year with low plant abundance had lower metabolic values: mean summer GPP was 22.3 g O₂ m⁻² d⁻¹ and mean summer ER was -15.1 g O₂ m⁻² d⁻¹. By the fall, GPP and ER were similar between the years. In the low flow years, mean summer metabolism was strongly net autotrophic but became less autotrophic over the season and into the fall. Conversely, the high flow year began slightly net heterotrophic and became strongly net autotrophic over the year. Compared to other plant-dominated streams, in the low flow years, the lower Yakima River had far higher GPP rates but similar or slightly higher ER rates. In the high flow year, GPP was higher, but less so, and ER was comparable. Overall, our results demonstrate that plants can have a strong effect on river ecosystem metabolism and that flow-driven plant removal can mediate this effect.