Soils of the Andean highlands are under threat from cropping system intensification. Improved forage-based fallows offer great promise to address this issue, but research is needed to better understand the potential of species mixtures vs. monocultures to support multiple farmer objectives, especially forage production and soil conservation. We used a pot study to quantify above- and belowground biomass production as well as the total N uptake of grass–legume pairs between five grasses: (1) oat (Avena sativa), (2) ryegrass (Lolium multiflorum), (3) festulolium (Lolium × Festuca genera), (4) brome grass (Bromus catharticus), and (5) orchard grass (Dactylis glomerata), and four legumes: (1) vetch (Vicia dasycarpa), (2) red clover (Trifolium pratense), (3) black medic (Medicago lupulina), and (4) alfalfa (Medicago sativa) relative to the performance of each species in monoculture within two soils from the central Peruvian Andes. Grass–legume bicultures demonstrated significant overyielding, producing 65% and 28% more total dry biomass and total N uptake on average than monocultures. Aboveground biomass of bicultures was significantly influenced by the species of legume present, while belowground biomass was more affected by the grass species in the mixture. When evaluating the growth of each species separately, our findings indicate that overyielding was driven more by the enhanced growth of grasses relative to legumes. Our findings indicate that combining key functional groups (e.g., grass and legume, annual and perennial) offers great promise for developing improved fallows for supporting soil health and productivity in Andean agroecosystems.