This thesis investigates model predictive control for distributed systems. To achieve a common task, all subsystems compute and share local predictions with their neighbors. Due to unreliable communication, shared data might be outdated, and uncertain predictions must be used for planning. The proposed control strategies explicitly account for both network-induced and local uncertainties, considering bounded disturbances in a robust control framework and stochastic models within a chance-constrained MPC setting. In addition, the strategies incorporate predictions of the communication network...