This research was concerned with the effects of the threat on the resistance to the dominating strategy in the interpersonal bargaining. 41 subjects participated in a Deutsch and Krauss's two-person bargaining game. In all conditions, a programmed confederate was used. He always took the shorter path and never com. promised. Four conditions were compared: in the equal condition (E), the usual version of the game was played. In the gate-no-use condition (GNU), only the confederate had a gate but never used it in the actual trials. In the gate-use condition (GU), only the confederate had a gate and closed it whenever the standoff occured on the one-way path. In the threat condition (T), only the confederate had a alternative choice. The gate could be used only when the standoff occured. If it was used, the trial would be terminated and both would then suffer the loss. But the magnitude of the loss was greater for the subject than for the confederate. In games, the gate was never used. The resistance was measured by the amount of time the subject spent attempting to go through the shorter path. The results were clear. In all conditions, subjects tended to respond with the resistance to yielding to the dominating confederate, but it decreased with time. The amount of it was greater in E and GNU than in GU and T. This meant that when there was the possibility that the resistance might lead to negative consequences, the tendency to resist decreased. The threat, even if it is used to force the other person to yield, can clarify the threatner's position and provide the clue to reach an agreement.