Eye-related bioelectric activity, such as blinks, eye movements, generates large amplitude electrooculography (EOG) artefacts negatively affecting the accuracy of electroencephalography (EEG) measurements. Using independent component analysis (ICA) for rejecting these artefacts also rejects some of the EEG information contained in the rejected component. This paper presents a novel method for the automatic identification and removal of the EOG artefacts to generate high quality EEG. The proposed method removes only the EOG activity and keeps the underlying EEG data unaltered.