<p>The purpose of this study was to develop a deeper understanding of the issues facing virtual team facilitators as they implement and facilitate virtual teams. The study asked the following research question: How do facilitators of virtual teams build relationships with their virtual team members? Because virtual teams are a new form of highly dynamic and ambiguous collaborative interaction, a major challenge of this study was the need to generate relevant data and analyze it in an appropriate manner. To achieve this, a research framework involving a training program format was instituted based on methods developed in Action Learning (AL), with data collection and analysis based on grounded theory approaches (Glaser and Strauss, 1967). The AL-based 'virtual team facilitation' training program used in this study was designed to achieve the following three goals: to generate interest and incentive for would-be participants, to give participants information and skills to initiate and facilitate their own virtual teams, and to generate data for analysis. After being recruited, participants were broadly interviewed to determine their prior experience with virtual teams and their perceived needs and concerns in implementing and facilitating their own virtual team. The researcher then developed a ten-week training program to meet these needs. A pilot program and two subsequent training programs were held. During the training programs, each participant planned for, or actually initiated and facilitated a virtual team within their own organizational context. Every two weeks the participants met with the researcher to investigate issues related to initiating and facilitating virtual teams and to discuss issues that were arising in their own virtual teams. In all seven participants from a variety of New Zealand organizations took part in the study. The seven participants formed a diverse group, from the managing director of a one-man, global virtual organization who worked exclusively in global virtual team settings to a self-employed consultant managing a local virtual work team. The participants were in various stages of their virtual team lifecycle, from planning through initiation to full-scale facilitation and evaluation of a just-completed virtual team project. The participants' virtual team project tasks ranged from managing a political campaign on the other side of the world to developing and running a national web-based academic assessment center. A unique feature of this study is that it involves organizational professionals as opposed to students. Data was collected from face-to-face and telephone interviews, group discussions and e-mail correspondences. Data collection extended to several months beyond the end of the training sessions. Using grounded theory techniques, the data was analyzed using "a general method of (constant) comparative analysis". Data was collected and coded simultaneously over the course of the training sessions, with subsequent coding confirming, refining, extending and modify the data. The data showed very clearly that the facilitators considered it essential to build some level of personal relationship with their virtual team members before commencing a virtual working relationship. Further, a unifying framework of three inter-related theoretical steps in the overall process a virtual facilitator goes through when building relationships with virtual team members was inductively derived from this study. These three steps are Assessing Conditions, Choosing Level of Relationship, and Creating Strategies. This study is the first to identify the steps a virtual team facilitator undertakes when building relationships with virtual team members. It has implications for virtual team practice, research and training.</p>