We consider a three-agent pursuit-evasion problem involving an attacker, and a target-defender team. The goal of the attacker is to capture the target, while that of the target is to escape the capture and the goal of the defender is to intercept the attacker before the attacker reaches the target. The defender and the target cooperate such that their objective is achieved efficiently. The target-defender team do not have accurate information of the attacker and hence use Extended Kalman Filter to estimate its states. The target-defender team uses Nonlinear Model Predictive Control (NMPC) to compute their optimal control inputs, while the attacker switches between pure pursuit and proportional navigation guidance laws about which the team is unaware of. The performance of the solution obtained from the proposed NMPC formulation was evaluated through numerical simulations and hardware experiments performed using ground rovers. © 2019 American Automatic Control Council.