In the light of shortening product life cycles and growing competition, supply chain efficiency has become a priority for researching pharmaceutical companies. Improving capacity management through dynamic product (re-)allocations is a widely acknowledged lever to achieve significant efficiency gains. However, product allocations in the pharmaceutical industry are subject to a specific source of uncertainty. Any facility that is to manufacture a given product for the very first time requires a production approval from regulatory authorities. The duration of the approval process is highly uncertain and can delay the completion of product allocations by several years. This implicitly affects decisions about outsourcing to contract manufacturing organizations (CMOs) who play an essential role in pharmaceutical supply chains (PSCs). Outsourcing is typically based on long-term contractual agreements and thus should be considered in a strategic time frame. To address these issues, we develop a two-stage stochastic MILP model for PSC network design under demand and approval time uncertainty. More specifically, we focus on the secondary manufacturing stage, which comprises all processes converting active pharmaceutical ingredients into finished drug products. A numerical study based on a real-life case of a global pharmaceutical company is presented to validate the applicability of our approach. The results show that approval times drive optimal allocation strategies and may induce different degrees of decentralization at upstream and downstream manufacturing stages. Furthermore, they are guiding make-or-buy decisions while uncertain demands mainly affect global capacity provisioning.