<p>Mixed crop-livestock systems (MCLS) are commonly perceived as a means of applying agroecological principles to achieve positive economic and environmental outcomes. However, a global meta-analysis on this claim is currently lacking. This paper contributes to filling this gap by gathering evidence from 700 studies worldwide, drawn from 270 publications spanning from 1992 to 2026. We find that 60% of the identified studies show that MCLS are economically and environmentally performant (double performance). Overall, the results show that MCLS decrease production costs (-23%), income variability (-15%), increase physical productivity (+ 5%), net income (+ 24%), reduce GHG emissions (-18%), non-renewable energy use (-37%), nitrogen surplus (-30%), pesticide use (-41%) and increase synthetic fertilizer use (+ 4%), biodiversity (+ 115%), and soil organic carbon (+ 23%). Among the studies showing that MCLS offer a double performance, almost 95% are based on experiments and simulations. However, when focusing on real farms, the results are very different, and the double performance is supported by only 12% of the studies. Results from real farms also show that MCLS can enhance the environmental performance of agricultural production systems, but they are not yet fully promising from an economic standpoint. The results also show that very few integration practices are tested in the studies and that these practices are often reported qualitatively. Meta-regression results confirm the potential of certain integration practices to convey economic and environmental benefits. They also highlight that conceptual definitions of MCLS should not be neglected in empirical analyses. Given the complexity of MCLS, further research, primarily based on real-world data and precise quantitative integration practices, is needed to identify optimal integration strategies for enhancing their actual benefits.</p>