The aim of our project is to examine the ability of people to learn how to use a computer system by exploration and to assess the efficacy of a user interface with facilities that are supposed to support exploration. In addition, the development of different types of knowledge that users have about the interface is tracked as they gain experience. We also examined the effect of individual learning style of the users on their exploratory behavior. In a first experiment it was found that subjects were indeed able to learn to use an electronic mail application that was unknown to them. No change in the user's knowledge about the interface could be measured, however, because the level of knowledge rose to a high level already in an early stage of the experiment. A second experiment did not show a significant advantage of the interface with the exploration-supportive facilities as compared to a version of the interface from which these facilities were removed or replaced by a simple paper user guide. In the third experiment a thinkaloud method was used to obtain more detailed information about the goals of the user and their realization. Here the exploration-support resulted in a more successful discovery and understanding of a number of the system's functions, in a better task performance in the second half of the series of tasks, and in a better procedural knowledge about the interface. The discrepancy with the findings of the previous experiment, in which subjects were not required to think aloud, is discussed. In this experiment, as well as in the previous experiment no effects of learning style were found.