BACKGROUND Intercropping of legumes with cereals enhances resilience, resource efficiency, and nutritional quality in food systems, yet few studies have integrated yield stability with grain quantity and post-harvest processability. Lentils (Lens culinaris Medik.) and hull-less barley (Hordeum vulgare L. var. nudum) represent a promising mixture for temperate climates. Here, we provide the first multi-year evaluation of protein content, protein yield stability, β-glucan concentration, and grain separation, across 3 years (2022–2024), two sites, and 22 treatments. RESULTS Mixed stands increased barley protein content by 18–26 g kg−1 dry matter (DM) compared with pure unfertilized and fertilized systems, whereas lentil protein content was marginally higher in pure stands. Mixed systems buffered environmental variability (Finlay–Wilkinson bᵢ ≈ 1.1), with variability substantially lower (CV = 38%) than in pure barley stands (63% CV), approaching that of pure lentil stability (29% CV). The lentil cultivar Beluga paired with the barley cultivar AF Cesar achieved the highest protein yield stability, recording a weighted absolute score of 94.5 via the AMMI (additive main effects and multiplicative interaction) yield stability index. β-glucan content (17–76 g kg−1 DM) was highest in fertilized pure AF Cesar and was influenced by year and site. Grain separation showed partially high impurity levels (>1%), due to similar seed sizes, but mixtures with contrasting seed sizes (e.g., Beluga–AF Cesar) improved separation (r = 0.38). CONCLUSION Intercropping of lentils with barley increases barley protein content and yield stability under Swiss conditions, providing the first evidence of effects on β-glucan content and seed-size-driven impurity reduction. Selecting cultivar pairs with contrasting seed sizes and optimizing sowing time can improve both agronomic and post-harvest performance in legume-cereal systems.