AbstractImportant drivers of gross primary production (GPP) and ecosystem respiration (ER) in lakes are temperature, nutrients, and light availability, which are predicted to be affected by climate change. Little is known about how these three factors jointly influence shallow lakes metabolism and metabolic status as net heterotrophic or autotrophic. We conducted a pan‐European standardized mesocosm experiment covering a temperature gradient from Sweden to Greece to test the differential temperature sensitivity of GPP and ER at two nutrient levels (mesotrophic or eutrophic) crossed with two water levels (1 m and 2 m) to simulate different light regimes. The findings from our experiment were compared with predictions made according the metabolic theory of ecology (MTE). GPP and ER were significantly higher in eutrophic mesocosms than in mesotrophic ones, and in shallow mesocosms compared to deep ones, while nutrient status and depth did not interact. The estimated temperature gains for ER of ~ 0.62 eV were comparable with those predicted by MTE. Temperature sensitivity for GPP was slightly higher than expected ~ 0.54 eV, but when corrected for daylight length, it was more consistent with predictions from MTE ~ 0.31 eV. The threshold temperature for the switch from autotrophy to heterotrophy was lower under mesotrophic (~ 11°C) than eutrophic conditions (~ 20°C). Therefore, despite a lack of significant temperature‐treatment interactions in driving metabolism, the mesocosm's nutrient level proved to be crucial for how much warming a system can tolerate before it switches from net autotrophy to net heterotrophy.