Open-loop controllability analysis plays an important role in integrated approaches to process design and control. Most existing open-loop controllability measures, such as right-half-plane zeros and the process resilience index, purport to measure the closed-loop performance but fail to indicate their overall effect on dynamic controllability. In this paper, based on the all-pass factorization, new controllability measures are proposed for linear stable multivariable processes. For a given process design, these controllability measures indicate the best achievable dynamic control performance subject to user-specified control constraints. Based on these measures, a controllability analysis approach is developed for selecting the best controllable process design from a variety of design alternatives.