Brain Computer Interfaces (BCI) represent a new communication option for those suffering from neuromuscular impairment that prevents them from using conventional augmented communication methods. This new approach has been developing quickly during the last few years, thanks to the increasing of computational power and the new algorithms for signal processing (Independent Component Analysis, Wavelets decomposition, Support Vector Machine etc.) that can be applied to the studies made on brain waves. Here follows two methodologies of approach based on making the computer adapt to the human brain activity and not vice-versa. The P300 and the SSVEP based BCIs, here presented, have the characteristics of not demanding specific training to the user.