In this study, multivariate analysis of the relationship between extruder barrel temperatures (100-140oC), initial feed moisture level (15-25%) and bambara groundnut flour (8-24%) content during twin-screw extrusion of rice-bambara groundnut blends for the production of complimentary food was investigated. Response surface methodology (RSM) and central composite experimental design (CCD) was used for the study. Twenty different blends were formulated based on CCD and extruded in a twin-screw extruder. Extruded samples were analyzed for nutritional quality using standard procedures and data fitted into a regression model. Results indicated significant variation among samples due to the process variability, with protein ranging between 16.99 and 27.36%, calorie value 391.14 and 397.99kcal/100g and Fe 10.1 and 13.02 mg/g sample. The fitted models for the prediction of linear, square and interaction relationships between the process and dependent variables were also significant (p≤0.05), adjusted coefficient of determinations (R2adjusted) > 90% and non-significant (p ≤ 0.05) lack-of-fit. Optimum process conditions of 120oC barrel temperature, 20g water/100g flour, and 22.4g bambara groundnut/100g flour were found to produce extruded complimentary foods having 21.66% protein, 394.33kcal/100g calorie, 1.34% fibre, 12.12mg/100g Fe and 25.81mg/100g Ca with a combined desirability function of 0.9102. The high protein of over 20% and 394.33kcal/100g calorie values are within the recommended values for normal growth and development of weaning and school children and therefore could be recommended for mitigating protein-energy malnutrition in developing countries, while the fitted models could be used in engineering modeling of extruder for food processing and new product development.