Although communication scholars usually have assumed that social interaction is intrinsic to the process of producing good group decisions, some writers have argued that social interaction actually may harm the quality of group decisions. Consequently, several noninteractive group decision-making procedures have been proposed; among these are the staticized group procedure and the nominal group procedure. While some research suggests the effectiveness of these non-interactive procedures, no research has compared directly the effect of staticized, nominal, and interacting decision procedures on the quality of group decisions reached with respect to complex, multistage tasks. The present study provides such a comparison. Reasoning that the open exchange of information, opinions, and criticism is necessary for optimal decision-making on complex tasks, it was hypothesized that groups employing an interacting procedure would produce better decisions than groups employing either staticized or nominal decision procedures. It was hypothesized further that subjects in the interacting condition would produce belter individual decisions than those in either the staticized or nominal conditions. Both of these hypotheses were confirmed. In addition, there was some evidence of an “assembly-effect bonus” among groups in the interacting condition.