Walking is a fundamental component of daily human activity, in which orderly execution of key gait events [such as heel strike (HS) and toe-off (TO) ] is essential for maintaining gait coordination and stability. However, the underlying brain–muscle neural coordination mechanisms associated with these events remain unclear. In this study, electroencephalography (EEG) signals from 19 channels and surface electromyography (sEMG) signals from 14 lower-limb muscles were synchronously recorded from 18 healthy participants during steady-state walking at a constant speed of 3.2 km/h. Brain–muscle connectivity networks were constructed using the adaptive directed transfer function (ADTF), and the dynamic connectivity characteristics associated with HS and TO events were quantitatively analyzed. The results showed that brain–muscle connectivity was strongest in the β frequency band. Intra-brain (EEG-EEG), brain-to-muscle (EEG-sEMG), and intra-muscle (sEMG-sEMG) connections were significantly stronger than muscle-to-brain (sEMG-EEG) connections. At both HS and TO events, information exchange between frontal-central cortical regions and the muscles of the supporting leg was markedly enhanced. Furthermore, compared with TO, the brain-muscle network at HS exhibited higher clustering coefficient, global efficiency, and betweenness centrality. These findings suggest that brain-muscle interactions during walking are predominantly mediated in the β band and dynamically modulated by gait events. Accurate characterization of gait event-related brain-muscle connectivity may provide important technical support for clarifying the neuromuscular coordination mechanisms underlying fine gait control.