This article addresses the fixed-time average consensus problem of nonlinear multiagent systems (MASs) subject to input delay, external disturbances, and switching topologies. Different from the finite-time convergence, the convergence time of the fixed-time convergence is independent of initial conditions. Then, an event-based control strategy is presented to reach the fixed-time average consensus under switching topologies and intermittent communication. Because the nonlinear dynamics, external disturbances, switching topologies, and triggering condition for intermittent communication are considered, the fixed-time consensus problem is more challenging under the event-based control than under the continuous-time control. Besides, a new measurement error is designed based on the hyperbolic tangent function to avoid Zeno behavior. Furthermore, an improved triggering function is designed to avoid continuous monitoring. Hence, resource consumption is reduced significantly. Finally, the effectiveness of the algorithms is validated by three simulation examples.