This paper considers both the distributed consensus control and output tracking problems of a group of unknown linear subsystems using the relative output information of neighboring subsystems. Each system is a minimum-phase SISO system with relative degree one and unknown parameters, and the interaction graph among the subsystems is directed. For the case where the directed graph is strongly connected, a distributed adaptive protocol is designed to achieve consensus. For the case where there exists an arbitrary constant reference signal, a distributed adaptive controller together with an internal model are presented to achieve output tracking in the sense that the subsystem outputs asymptotically follow a reference constant. The proposed adaptive protocols are independent of the parameters of the subsystems, only use the relative outputs of neighboring subsystems, and hence are fully distributed. (C) 2015 Elsevier Ltd. All rights reserved.