<p>This paper investigates the hierarchical distributed data-driven formation tracking control problem for heterogeneous multi-agent systems (HMASs) subject to external disturbances. The HMASs consist of multiple unmanned aerial vehicles (UAVs) and multiple unmanned ground vehicles (UGVs). Firstly, the HMASs are transformed into an equivalent linear data model using the compact form dynamic linearization (CFDL) technique, and an integrated model is then established for the theoretical analysis. Then, a hierarchical distributed data-driven formation tracking control protocol is constructed for all follower UAVs and UGVs, allowing them to track the virtual leader’s trajectory and achieve a predefined formation shape. Considering that the leader’s state cannot be directly accessed by all followers, an adaptive distributed observer is firstly developed for the state estimation. Then, an adaptive extended state observer (AESO) is designed for the real-time estimation of external disturbance signals. Subsequently, a data-driven adaptive tracking controller is designed with disturbance rejection, and the stability of the heterogeneous UAVs-UGVs is proved utilizing the linear matrix inequality technique and Lyapunov stability theory. Finally, simulations are provided to verify the effectiveness of the proposed control algorithm.</p>