A noninvasive brain-computer interface (BCI) using scalp-recorded electroencephalographic (EEG) activity and an adaptive algorithm enables humans, including those with spinal cord injuries, to control a robotic arm or neuroprosthesis with multidimensional point-to-point movement. This study demonstrates that noninvasive BCIs can achieve performance comparable to invasive BCIs in terms of movement time, precision, and accuracy. The adaptive algorithm identifies and focuses on EEG features the user can best control, improving performance over time. The results suggest that people with severe motor disabilities could use brain signals to operate robotic arms or neuroprostheses without needing implanted electrodes. The study involved four users with varying levels of BCI experience. The BCI system used sensorimotor rhythms recorded from the scalp to control cursor movement in two dimensions. The adaptive algorithm adjusted weights to optimize the translation of EEG signals into cursor control. Users were able to move the cursor to targets with high accuracy, and the system showed independence between vertical and horizontal movements. The study also found that users could control the cursor to novel locations with only a slight increase in movement time, indicating the system's adaptability. Compared to previous noninvasive studies, the current study achieved significantly higher correlations between EEG signals and target locations, with minimal correlation with the wrong dimension. This indicates more accurate and independent control. The study also compared results with invasive studies in nonhuman primates, finding that the noninvasive BCI's performance falls within the range reported for invasive methods. The results suggest that noninvasive BCIs could support clinically useful operation of robotic arms, motorized wheelchairs, or neuroprostheses. Potential improvements include expanding the adaptive algorithm to include additional EEG recording locations, frequency bands, and time-domain features, refining user training protocols, and improving the translation of EEG features into cursor movements. Recent studies suggest that using cortical surface recordings (electrocorticographic activity) could further enhance BCI performance, offering a minimally invasive alternative to implanted electrodes. The study concludes that noninvasive BCIs can provide real-time multidimensional movement control, potentially eliminating the need for brain implants. This could make BCIs an important communication and control option for people with severe motor disabilities. The research was supported by various grants and was conducted with the assistance of multiple collaborators.A noninvasive brain-computer interface (BCI) using scalp-recorded electroencephalographic (EEG) activity and an adaptive algorithm enables humans, including those with spinal cord injuries, to control a robotic arm or neuroprosthesis with multidimensional point-to-point movement. This study demonstrates that noninvasive BCIs can achieve performance comparable to invasive BCIs in terms of movement time, precision, and accuracy. The adaptive algorithm identifies and focuses on EEG features the user can best control, improving performance over time. The results suggest that people with severe motor disabilities could use brain signals to operate robotic arms or neuroprostheses without needing implanted electrodes. The study involved four users with varying levels of BCI experience. The BCI system used sensorimotor rhythms recorded from the scalp to control cursor movement in two dimensions. The adaptive algorithm adjusted weights to optimize the translation of EEG signals into cursor control. Users were able to move the cursor to targets with high accuracy, and the system showed independence between vertical and horizontal movements. The study also found that users could control the cursor to novel locations with only a slight increase in movement time, indicating the system's adaptability. Compared to previous noninvasive studies, the current study achieved significantly higher correlations between EEG signals and target locations, with minimal correlation with the wrong dimension. This indicates more accurate and independent control. The study also compared results with invasive studies in nonhuman primates, finding that the noninvasive BCI's performance falls within the range reported for invasive methods. The results suggest that noninvasive BCIs could support clinically useful operation of robotic arms, motorized wheelchairs, or neuroprostheses. Potential improvements include expanding the adaptive algorithm to include additional EEG recording locations, frequency bands, and time-domain features, refining user training protocols, and improving the translation of EEG features into cursor movements. Recent studies suggest that using cortical surface recordings (electrocorticographic activity) could further enhance BCI performance, offering a minimally invasive alternative to implanted electrodes. The study concludes that noninvasive BCIs can provide real-time multidimensional movement control, potentially eliminating the need for brain implants. This could make BCIs an important communication and control option for people with severe motor disabilities. The research was supported by various grants and was conducted with the assistance of multiple collaborators.