The thesis presents quantitative decision support models to improve managerial decision making processes for optimal allocation of raw milk to end-products and production planning of products (so-called milk valorization). The complexity of these decisions is driven by numerous constraints such as market demand, production capacity, interrelated production processes. The developed methods are applied to the case of FrieslandCampina, which confirmed its applicability and contribution to successful valorization of milk. Results show that a comprehensive dairy valorization model is needed to successfully valorize milk. Production recipes should be based on the components of milk instead of milk volumes. Furthermore, the research showed the potential value of by-product valorization and developed a method to determine the robustness (stability) of valorization plans. Finally, a framework is presented for performance assessment of processing units allowing identification of inefficient units and potential performance improvement opportunities.