In this paper we extend the framework of threat bargaining games to include those with incomplete information. In this set up we address ourselves to two significant problems: (1) Under what conditions would 'truthful' revelation of the disagreement payoffs by a Nash equilibrium of the threat bargaining game? (2) Obtaining a characterization of the Nash bargaining solution without the Independence of Irrelevant Alternatives Assumption. Our framework of analysis is general enough to include within its purview the study of non-cooperative bargaining problems with incomplete information, played by Bayesian players, although the specific problem addressed to in this paper does not fall in that category.