This study examined the utility of the conceot of interaction tendency in explaining personal spacing behavior. Interaction tendency v/as defined as an aggregate of feelings about an interaction situation and as a mediator between personal spacing behavior and the kind of associations connected with an interaction. It was hypothesized that as the positivity of the associations connected with an interaction increased, interaction tendency increased, and personal spacing decreased. Two levels of task (problem solving and conversation), two levels of associations connected with the task (positive and negative), and tv;o levels of associations connected with the other interactant (positive and negative) were manipulated, and the resultant seating behavior was observed. It was found that the independent variables were not predictive of seating behavior.