The development of brain/neural-controlled exoskeletons allow for restoration of movements in paralysis. By translating brain activity associated with the intention to move, such systems enabled, e.g., quadriplegic patients with complete finger paralysis to eat and drink in an outside restaurant. However, noninvasive means to record brain activity often lack sufficient signal quality for reliable and safe operation, particularly in noisy, uncontrolled environments or presence of muscle artifacts due to whole body movements. Thus, hybrid control paradigms were developed that merge different biosignals to increase reliability of exoskeleton control. Here, we introduce such control paradigm for restoration of gait using a personalized exoskeleton based on electroencephalographic and electrooculographic (EEG/EOG) signals. While exoskeleton movements were initiated by event-related desynchronization (ERD) of sensorimotor rhythms (SMR) associated with the intention to walk, the exoskeleton was stopped by a specific EOG signal. Using such paradigm does not only provide intuitive control, but may also trigger neural recovery when used repeatedly over a longer period of time. Further validation of this approach in a larger clinical study on gait assistance and rehabilitation will be needed.