Common airframe design and certification processes are believed to be too expensive, lengthy, and inflexible. This difficult situation is exacerbated for military airframes by the current push toward nontraditional designs that offer unique performance capabilities or will be required to operate in extreme environments, and therefore may require equally unique test programs. Consistent and early use of high-fidelity computational tools is being looked to as a primary means of overcoming the inadequacies of current processes, especially through the reduction or acceleration of the required test regimen. Uncertainty quantification methods are concurrently being pursued to facilitate the validation of computer models and, more generally, to facilitate improved decision making through methodical risk assessment. Robust use of these methods in an integrated design process could require substantial modification of existing systems engineering frameworks and funding profiles. We discuss risk allocation in airframe systems engineering. Particular attention is given to the concept of allocating system-level risks in multidisciplinary design problems, such as the avoidance of aeroelastic instabilities. Finally, potential methods are discussed for improving the storage and retrieval of design information that is crucial to accurate risk assessment