Reducing vulnerability and enhancing resilience of infrastructure networks subject to uncertain disruptions is a challenging task as these networks become increasingly interconnected and interdependent. This article devotes to addressing the preparedness planning problem of interdependent infrastructure networks. The interdependency between infrastructure networks is characterized by a two-way physical interdependency, that is, the state of one infrastructure network is dependent on that of another infrastructure network, and vice versa. Furthermore, the multistate characteristics of infrastructure networks are taken into account and accommodated to the studied physical interdependency. A tailored two-stage stochastic programming framework, which is able to cope with the uncertainty associated with disruption scenarios, is put forth to facilitate an effective resilience enhancement strategy under a limited budget. In this framework, the first-stage problem selects the optimal protection levels of network components in advance, whereas the second-stage problem optimizes the network operation after a disruption scenario occurs. The proposed framework is implemented to an illustrative system composed of interdependent power–water distribution networks to demonstrate the effectiveness of the resilience enhancement strategy.