<p>This paper addresses the fixed-time consensus problem for first-order multi-agent systems under stochastic switching topology. A distributed fixed-time consensus protocol for the first-order multi-agent systems is given by modeling the switching topology as an ergodic Markov process. The proposed approach only requires the union of time-varying topologies to be connected, which significantly relaxes the requirements on communication topologies. Additionally, the backstepping method is utilized alongside the system’s dynamic to design a state observer capable of estimating agent’s states. Based on the observer, a novel control protocol is proposed. The introduction of the observer effectively mitigates the impact of disconnected topology on system convergence, and the stability of the system is improved. The effectiveness of the observer is rigorously analyzed using algebraic graph theory and Lyapunov stability theory. Notably, once the observer error converges with the Markov switching process, the subsequent convergence time of the system is no longer be affected by topology switches, ensuring fixed-time consensus. Finally, the effectiveness of the algorithm is verified through numerical case studies.</p>