We often hear that global knowledge work teams are affected by time zone differences, but most research in geographically dispersed collaboration has focused on the effects of distance and has treated time zones as a secondary factor. The experimental study we describe here is part of a larger research program aimed at understanding how technical teams coordinate their work across time zones and which factors influence performance. In this study we investigate how time zone differences affect team performance in a laboratory setting. The study is composed of three phases corresponding to different task type manipulations – simple, complex and equivocal task. In each of the study phases, dyadic teams were randomly assigned into 4 time zone (i.e., work time overlap) conditions: full overlap, 2/3 overlap, 1/3 overlap and no overlap. Teams performed a map drawing task simulating the assembly of software components. We completed data collection for 131 dyad teams. In each phase we collected the following data: team performance (speed and accuracy), exit survey, and chat log capture. In this paper we describe our research design, briefly discuss our preliminary results from analysis of data from Phase 1, and describe our expectations and next steps for the full study. In Phase 1 we found that time separation has a negative effect on accuracy. We also found that a small amount of time separation has a negative effect on production speed but, surprisingly, speed actually increases with further increases in time separation – a “U” shaped curve. Our chat log text analysis also revealed differences in communication patterns across time zone conditions, which helps explain the unanticipated results. To evaluate if the simplicity of our task influenced our results in Phase 1 we manipulated the task in Phase 2 (added complexity) and Phase 3 (more equivocal). Expected results and implications from these subsequent phases are discussed at the end.