This paper presents a new iterative learning design approach for cooperative tracking and estimation of linear multi-agent systems with a dynamic leader. The input of the dynamic leader is unavailable to all follower agents. Distributed iterative lea