For several disabled people a system is needed, that provides a new way to facilitate interaction with their surroundings. A Brain-Computer Interface (BCI) is a non-muscular channel that allows motor disabled people to drive a computer directly by their brain activity. To date, motor imagery tasks (i.e. imagined movement) are most commonly used to control a BCI, but to address motor disabled people and their individual histories, a non-motor imagery based BCI could help to provide a new communication channel for people with individual motor disabilities since years or from birth on. Hence, a four-class BCI driven by the mental tasks Word association, Mental subtraction, Spatial navigation and Motor imagery, was implemented. As the P3b component is known to be associated with task relevant processes, the study focuses on posterior regions.The sample consisted of twelve participants, who attended two Screening and eight Feedback sessions, trying to drive a BCI by mental imagery. Additionally, auditory stimuli were presented using an oddball paradigm including five standard tones and one target tone, whereby in the active condition the users had to respond to the target tones by pressing a button, and in the passive condition they were instructed to ignore all of the presented tones. The recorded data was analyzed for the P300 components, namely amplitude (V) and latency (ms), assuming, that changes in task difficulty are reflected in those components and lead to evidence concerning workload each task produces and the user?s remaining cognitive capacities. The results suggest that the four mental tasks can be discriminated according to the workload each mental task produces. These findings signify that the P300 component is a reliable tool to measure the workload each mental task produces and the user?s remaining cognitive resources as well as to illustrate the differences in mental tasks regarding their task difficulty.