Musicians and composers have been using brainwaves as generative sources in music for at least 40 years and the possibility of a brain-computer interface for direct communication and control was first seriously investigated in the early 1970s. Work has been done by many artists and technologists in the intervening years to attempt to control music systems with brainwaves and — indeed — many other biological signals. Despite the richness of EEG, fMRI and other data which can be read from the human brain, there has up to now been only limited success in translating the complex encephalographic data into satisfactory musical results. We are currently pursuing research which we believe will lead to the possibility of direct brain-computer interfaces for rich and expressive musical control. This report will outline the directions of our current research and results.