Abstract Human activity has induced a multitude of global changes that are likely to affect the functioning of ecosystems. Although these changes act in concert, studies on interactive effects are scarce. Here, we conducted a laboratory microcosm experiment to explore the impacts of temperature (9, 12 and 15 °C), changes in soil humidity (moist, dry) and plant diversity (1, 4, 16 species) on soil microbial activity and litter decomposition. We found that changes in litter decomposition did not mirror impacts on microbial measures indicating that the duration of the experiment (22 weeks) may not have been sufficient to determine the full magnitude of global change effects. However and notably, changes in temperature, humidity and plant litter diversity/composition affected in a non-additive way the microbial parameters investigated. For instance, microbial metabolic efficiency increased with plant diversity in the high moisture treatment but remained unaffected in low moisture treatment suggesting that climate changes may mask beneficial effects of biodiversity on ecosystem functioning. Moreover, litter decomposition was unaffected by plant litter diversity/composition but increased with increasing temperature in the high moisture treatment, and decreased with increasing temperature in the low moisture treatment. We conclude that it is inevitable to perform complex experiments considering multiple global change agents in order to realistically predict future changes in ecosystem functioning. Non-additive interactions highlight the context-dependency of impacts of single global change agents.