Sown biodiverse permanent pastures rich in legumes (SBPPRL) were developed in Portugal in the 1960s and 1970s as a strategy to increase grassland productivity by sowing mixtures of up to 20 species/cultivars of legumes and grasses. Compared to semi-natural pastures, the resulting engineered system provides higher yields of better quality pasture, significantly increasing sustainable stocking rates, with multiple environmental co-benefits. Here, we propose a conceptual framework for the sustainability assessment of SBPPRL and apply it with existing data. Our objective is to inquire if this system is an example of sustainable intensification of livestock production, i.e., an economic and ecological win–win solution that can answer many of the causes for ecosystem degradation in semi-arid and sub-humid climate zones, such as in the Mediterranean basin. We build on experimental results from previous studies, which suggest that SBPPRL replenish soil organic matter pools and improve soil structure. The high increase in stable soil organic matter acts as a carbon sink, turning the system into an optimum tool for climate change mitigation and adaptation. Portugal made use of this fact by supporting the expansion of SBPPRL areas and abating the corresponding carbon from Kyoto Protocol emissions calculations. We resorted to the literature to evaluate other environmental effects due to the absence of data specifically for SBPPRL. Surface water runoff decreases and pirophyte shrub vegetation is eliminated or much reduced. Nitrogen accumulates in stable forms in the soil after being fixed by Rhizobium/legume symbiotic associations. Legumes depend on phosphorus fertilization; as such the nitrogen cycle in SBPPRL relies on a potentially non-renewable resource (required during the first years after installation of the pasture), which may be a potential limiting factor in the future. The effects on wild biodiversity are unclear. The methodology laid out in this article provides an innovative framework to assess these effects as additional experimental data becomes available.