Decomposition plays a key role in carbon cycling, as it determines how long carbon remains stored in recalcitrant organic matter, before it is converted into more labile forms like CO2. Changes in temperature and precipitation due to climate change have the potential to alter ecosystem processes such as decomposition. Predicting the effects of climate change on decomposition rates in freshwater streams could be a critical step toward elucidating how organic matter availability will change under future climate scenarios. This study used the global CELLDEX decomposition data set and an existing boosted regression tree (BRT) model to predict stream decomposition rate at watershed scales under two climate scenarios: shared socioeconomic pathway 126 (SSP126) and shared socioeconomic pathway 585 (SSP585). Three climate variables extracted from Earth System Models (ESMs) predicting conditions in 2100 were used as inputs in the BRT model: mean annual air temperature, temperature range and mean annual precipitation. We compared predicted changes in decomposition rates in equatorial Lake Victoria and temperate Lake Superior watersheds, as well as across smaller watersheds in five different biomes from polar to equatorial. A sensitivity analysis was conducted to determine which climate variable(s) had the greatest impact on decomposition rate in each watershed. There was a general trend of increased decomposition rates under both shared socioeconomic pathways in all but one of the watersheds studied. However, the sensitivity analysis showed that the changes in decomposition rate and the degree of impact of each climate variable were regionally specific. The combined change in temperature and temperature range explained most of the variation in decomposition rates. However, precipitation change was the most important climate variable in the boreal Agashashok River watershed. Additionally, temperature change was the most important in the Lake Superior watershed, possibly due to maritime influence from the lake causing sustained periods of high temperatures. The results of this study highlight the sensitivity of a key ecosystem process to climate change and provide biome-specific guidance on important climate drivers.