Changing market conditions, volatile customer demand, intense competition and tightness of capital are some of the primary characteristics of the global economy that affect process industries nowadays. The main objective of the thesis is to facilitate business decision-making in today's increasingly complex and highly uncertain market environment by applying mathematical programming techniques for multi-site capacity planning and business optimisation in process industries. In the first part of the thesis, the problem of multi-site capacity planning under uncertainty in the pharmaceutical industry is addressed. A comprehensive two-stage, multi-scenario mixed-integer linear programming (MILP) model is proposed able to determine an optimal product portfolio and multi-site investment plan in the face of clinical trials uncertainty. A hierarchical algorithm is also developed in order to reduce the computational effort needed for the solution of the resulting large-scale MILP model. The applicability of the proposed solution methodology is demonstrated by a number of illustrative examples. The second part addresses the problem of business optimisation for customer demand management in process industries. A customer demand forecasting approach is developed based on support vector regression analysis. The proposed three-step algorithm is able to extract the underlying customer demand patterns from historical sales data and derive an accurate forecast as demonstrated through a number of illustrative examples. An active demand management approach for close substitute products is also developed based on price optimisation. The proposed methodology is able to determine optimal pricing policies as well as output levels, while taking into consideration manufacturing costs, resource availability, customer demand elasticity, outsourcing and market competition. An iterative algorithm is developed able to determine Nash equilibrium in prices for competing companies as demonstrated by the illustrative examples.