This dataset comprises EEG recordings from 84 healthy participants performing motor imagery tasks across multiple controlled environments (laboratory and simulated hospital), with a supplementary cohort of 3 participants recorded in microgravity conditions in a space station environment. The dataset includes multiple motor imagery paradigms (Graz, SSMVEP-MI, and hybrid variants) with 64-channel EEG recordings at 1000 Hz sampling rate (69 channels acquired, 60 channels used in analysis), designed to investigate environmental effects on brain-computer interface performance and enable cross-environment robustness evaluation. This dataset is a republication of OpenNeuro ds007221 on the NEMAR platform to enhance accessibility and discoverability within the neuroimaging research community.