The U. S. Army is developing distributed interactive simulation (DIS) systems for combat training and military concept development, testing, and evaluation. The early emphasis and implementation has been on linking vehicle simulators, without providing for the training or participation of dismounted soldiers (Knerr, et al., 1994). The Army Research Institute for the Behavioral and Social Sciences (ARI) and Defence Research and Development Canada are investigating distributed training for dismounted soldiers. Unlike vehicle simulators, where the crew are able to interact within the simulator unimpeded by the simulation technology, simulators for dismounted combatants interpose the limitations of simulation technology between team members. Any effect this has on the acquisition of team skills is exacerbated in distributed simulations because interaction between team members will be further limited. To investigate the nature and severity of this situation, the reported experiment addresses the development of team coordination under conditions of either distributed or local mission rehearsal. In the virtual environment (VE) scenario two person teams search buildings for hazardous materials and neutralize them while being opposed by computer generated forces. Each participant is trained to standard criteria on all tasks and activities before the team is formed and mission rehearsals begin. Each team then performs eight missions with an after-action review (AAR) after each mission. Mission sessions are distributed over several days. The preliminary data show improvement in team overall performance in the number of rooms searched, time to perform collective tasks, and hazardous materials disarmed. We anticipate that teams trained using distributed simulation, having more limited opportunities to interact away from the mission and AAR, will not develop the same levels of performance achieved by those teams trained in the same location.