The boundary role conflict suggests two types of functions in negotiations: monitoring the other side for evidence of movement and monitoring one's own side for evidence of preferences. These functions differ in terms of focus and information-processing. This paper addresses these functions in terms of two general models, referred to as the negotiator as bargainer and the negotiator as representative. The negotiator-as-bargainer model assumes responsiveness between opposite-number negotiators. Two versions of this model showed that responsiveness can be based on both one's own previous concessions and the other's concessions or it can be a more complex function of expectations and evaluations. Each of the versions was supported in part by the data which suggested that these may be early and late processes in negotiation. The negotiator-as-representative model assumes responsiveness between the negotiator and his constituents. Such responsiveness is depicted in the form of a utility model where the negotiator attempts to balance n-components of value in the process of building a package. He is concerned with maximizing the value of the package in terms of both his own and his constituents' priorities. Experimental results suggest that the model accounts for a significant portion of variance in actual decisions. Finally, implications are drawn toward a reconceptualization of the boundary role conflict.