This paper proposes a new cooperative guidance law for intercepting a maneuvering target based on an interacting multiple model to address the problems of unknown target trajectories and insufficient accuracy of cooperative guidance. First, based on the concept of time-to-go inversion guidance (TIG), an explicit guidance law which can give accurate time-to-go solutions is designed, and the cooperative guidance law for stationary targets are achieved by introducing a biased term. Then, considering the circumstances of the unknown maneuvering target, the position of the target is tracked accurately by using the concept of interactive multiple models (IMMs). Based on the tracking results, the future trajectories of the target can be predicted accurately by using autoregressive integrated moving average models (ARIMA). Furthermore, the proposed cooperative guidance law and proposed prediction method are incorporated to construct a cooperative guidance method for intercepting maneuvering target. Numerical simulations demonstrate that the proposed guidance law can intercept maneuvering targets successfully even though in the circumstances of large initial heading errors or unknown maneuvering targets, which validate the effectiveness of the proposed cooperative guidance framework against maneuvering targets and the accuracy of the proposed prediction method in predicting target maneuvering intentions.