Falling is one of the major causes of unintentional death or injury worldwide. People older than 65 years and people suffering from neurological disease have a higher chance of falling, which indicates the involvement of higher brain structures such as the cortex. For this reason, cortical activity underlying balance control has been subject of research using mobile neuroimaging methods during balance tasks. Often, the focus is on the moment balance is lost and a compensatory action is made. In this study, differences between two different strategies are investigated. Healthy, young participants receive balance perturbations by backward translation of a platform on which they stand, while recording high-density electroencephalography (EEG) to study task-related brain activity. EEG is the recording of changes of electric potential related to brain activity. Participants receive two instructions; in the first condition, they are instructed to respond naturally, which means they take a step. In the second condition, participants should try to inhibit their step. It is hypothesised that the second condition will reflect neural responses with a higher magnitude, because it is more difficult to respond by step inhibition. The dynamic nature of the task introduces a muscular and motion artefacts to which the EEG is sensitive. A blind-source separation algorithm (independent component analysis) in combination with an automatic outlier rejection algorithm is used to separate artefacts from neural sources. The obtained neural sources are divided into clusters based on their location on the scalp, dipole representation, power spectrum and timecourse. The resulting clusters represent the motor cortex (legs, left and right hand area), the prefrontal cortex, and the temporal-occipital areas for the left and right hemisphere. Analysis in time-domain shows a negative peak in the timecourse directly after the perturbation. In previous research, this peak has been related to error detection, which in this experiment is the loss of balance. In time-frequency domain, at the motor cortex, phase synchronisation is found at 5-25 Hz, followed by phase desynchronisation in the 8-15 Hz and 20-30 Hz frequency range. This is also in accordance with previous literature. However, there was no significant difference between the two conditions. The hypothesis is therefore rejected, at this point. One explanation to this result may be the small dataset used for this experiment. During preprocessing, up to 65% of data was rejected for some participants. An increase in the number of trials and the duration of delays in between trials, may improve the quality of the dataset such that the hypothesis can be reconsidered. Nonetheless, the correspondence in results with previous research already indicate that the motion platform is suitable for EEG experiments.