Optimizing the structure and development pathway of biopharmaceutical drug portfolios are core concerns to the developer that come with several attached complexities. These include strategic decisions for the choice of drugs, the scheduling of critical activities, and the possible involvement of third parties for development and manufacturing at various stages for each drug. Additional complexities that must be considered include the impact of making such decisions in an uncertain environment. Presented here is the development of a stochastic multi-objective optimization framework designed to address these issues. The framework harnesses the ability of Bayesian networks to characterize the probabilistic structure of superior decisions via machine learning and evolve them to multi-objective optimality. Case studies that entailed three- and five-drug portfolios alongside a range of cash flow constraints were constructed to derive insight from the framework where results demonstrate that a variety of options exist for formulating nondominated strategies in the objective space considered, giving the manufacturer a range of pursuable options. In all cases limitations on cash flow reduce the potential for generating profits for a given probability of success. For the sizes of portfolio considered, results suggest that naively applying strategies optimal for a particular size of portfolio to a portfolio of another size is inappropriate. For the five-drug portfolio the most preferred means for development across the set of optimized strategies is to fully integrate development and commercial activities in-house. For the three-drug portfolio, the preferred means of development involves a mixture of in-house, outsourced, and partnered activities. Also, the size of the portfolio appears to have a larger impact on strategy and the quality of objectives than the magnitude of cash flow constraint.