This paper is concerned with the problem of model reference tracking control for a class of linear networked control systems (NCSs) in which a controlled plant is connected to an observer-based controller via a communication network. In the presence of network-induced delays and packet dropouts in the sensor-to-controller and controller-to-actuator connections, the inputs of the plant and the controller are updated in an asynchronous way. In this case, the resulting NCS is equivalent to a linear system with two interval time-varying delays. A sufficient stability condition that ensures the NCS with an H∞ tracking performance is derived by using a Lyapunov-Krasovskii functional approach. Due to the asynchronous input errors, a separation principle cannot be used to design the observer-based controller. A novel design algorithm of tracking control is presented by using the stability condition and a particle swarm optimization (PSO) technique. The effectiveness of the algorithm is illustrated by a numerical example.