Water temperature is recognized as an important driver of aquatic ecosystems and is of increasing interest because of global warming. However, to date, no studies have disentangled the contribution of annual thermal variability from the contributions of annual thermal mean and other environmental variables to spatiotemporal patterns of macroinvertebrate assemblages in mountain streams. In this study, we explored patterns of macroinvertebrate biodiversity within and among 5 mountain stream sites in the Serio watershed (Orobic Alps, Northern Italy) along a gradient of annual water thermal variability (annual range: ~1–16°C) but with similar means. At each site, macroinvertebrates were sampled; physicochemical variables, flood disturbance, and resource availability were measured monthly over a year; and water temperature was monitored continuously. The relationship between macroinvertebrate assemblages and environmental conditions was described by separating time and space through co-inertia analyses. Results showed a strong temporal co-inertia, with floods and mean monthly temperature most associated with within-site macroinvertebrate temporal patterns (70% of total variance) and with the greatest effects on the abundance of stoneflies. Synchronous floods occurred in all sites and promoted changes in macroinvertebrate community composition through a loss of taxa and individuals, whereas monthly water temperature variations were associated with the replacement of individuals among taxa. Spatial differences (30% of the total variance) were mainly linked to geographical proximity, although annual thermal variability covaried with annual temporal dissimilarity in taxonomic composition, suggesting that temporal patterns also explain spatial differences. Overall, the study indicates that even when accounting for the overall environmental setting, the thermal regime remains an important driver of macroinvertebrate spatiotemporal patterns. Mountain streams are particularly subject to thermal variation, and an increase in mean water temperature and thermal variability will substantially influence community composition, with the strongest effects on cold stenothermal taxa. Large-scale research across mountain ranges is essential to properly assess how spatiotemporal thermal heterogeneity influences the distribution of macroinvertebrates, as well as to guide conservation strategies in the context of increasing global warming.