Virtual Reality (VR) have experienced a resurgence in recent years with the introduction of the Oculus Rift, and this new market have made the technology widely available for consumers. With this renewed popular interest in VR, there has also been an increase in research around the field. While this research is introducing inventive and creative new hardware and techniques, the field still lacks an underlying foundation of an empirical and conceptual character. This problem also extends to the particular field of text input in VR using virtual keyboards, which also have the same issues. To move this field forward, this thesis will identify and examine several common techniques, and evaluate the user experience and usability of these. We ended up implementing and evaluating four different virtual keyboards in an empirical comparative study, where we gathered qualitative data through interviews, observation and questionnaires, and also quantitative data from logging the participants performance. We used semi-structured interviews in combination with written responses in questionnaires to form the basis of the qualitative data, which were analyzed using open coding. To measure user experience and usability, we used the Game Experience Questionnaire and System Usability Scale, which are both regarded as robust for the field, and which we used to validate the interviews and written responses. The performance data was also analyzed with respect to the qualitative data. The results showed significant differences in most measurements between the techniques, and we used this to validate the qualitative data we obtained. The main results that we obtained from this implied that virtual keyboards should embrace all modalities enabled by utilizing the immersive qualities of VR, which are highly valued by the users. Also, users by far prefer fast, immersive and enjoyable techniques over precise, but slow techniques.