Distributed computing system consists of number of independent computers located at different locations. These computers may become faulty or unresponsive at any point of time. Such failures will affect the overall performance of the distributed system. Thus, it becomes necessary to detect such faulty computers or nodes. This paper proposes a new adaptive fault diagnosis algorithm namely, Coordinator-based Adaptive Fault Diagnosis algorithm for distributed systems. It uses an innovative technique where a coordinator pair supervises and detects the faulty nodes in the system. This is an adaptive fault diagnosis algorithm for t-diagnosable system, where ‘t’ is the upper bound for the number of detectable faulty nodes in the system. This algorithm also allows re-entry of repaired faulty nodes in the next diagnostic cycle. This algorithm is periodically executed at every node to detect the faulty nodes in the distributed computer network. The implementation of the proposed algorithm on a small laboratory setup is presented in this paper along with the results.