This dissertation discusses the primary features of an electroencephalographic (EEG) human-computer interface. The paper begins with an overview of the human-computer interface problem and covers issues such as safety, cost, and complexity of technology. A detailed description of the equipment used throughout the research is given. This description includes the electrodes and their placement, the neurodata amplifier, the analog-to-digital converter, the analysis software, and the host computer. The foundation for realizing an electroencephalographic human-computer interface is based primarily on two possible paradigms. The first is a direct thought-controlled interface in which an attempt is made to correlate the EEG data with conscious thoughts of the subjects. Topics discussed in this area include raw-data analysis, chaos theory, autoregressive modelling, biofeedback and asymmetry spectral analysis. The second and most promising method for achieving an interface involves the use of visual evoked potentials. The primary evoked- potential issues presented are verification of the driving response, frequency separation, transient analysis and object multiplicity.