Uncertainty in new product development (NPD) planning embraces market, creative, technological, and process dimensions to a much greater extent than in non-NPD project planning. Yet, NPD management is becoming increasingly decentralized, both within the firm and across the supply chain. Hence, planning for NPD uncertainty often results in path-dependent scenarios cutting across the strategic, tactical, and operational levels of planning. To coordinate this resulting complexity, we propose a stochastic hierarchical product development planning framework with multiple recourses, i. e., corrective actions, to maximize performance across a firm's entire NPD program. We also argue the necessity for a fourth planning level, the infrastructural, that reestablishes norms for market projections, technological forecasts, scheduling, and requirements as latent uncertainty in the environment is continually revealed. An illustration from the automotive industry is presented to demonstrate a deployment of our framework. We additionally discuss the applicability of this framework for managing NPD capabilities over time.