This paper considers a market where pairs of agents who are interested in carrying out a transaction are brought together by a stochastic process and, upon meeting, initiate a bargaining process over the terms of the transaction. The basic bargaining problem is treated with the strategic approach. The paper derives the steady state equilibrium agreements; analyzes their dependence on market conditions such as the relative numbers of agents of different types; and discusses their relations with the competitive equilibrium outcome and other results in the search equilibrium literature. THIS PAPER CONSIDERS a market that operates in the following way. Pairs of agents who have mutual interest in carrying out a transaction are brought together by a stochastic process. When two agents meet, they initiate a bargaining process over the terms of the transaction. If two agents reach an agreement a transaction takes place and they leave the market. Of course, the bargaining positions and hence the agreement reached in any particular meeting will be affected by the conditions prevailing in the market. These will include the chances that each of the negotiating parties have of meeting other partners in the event that the agreement in the current negotiations is delayed; also the expected length of time required to achieve an alternative transaction, and the expected behavior of alternative partners. The study of such a market mechanism is of interest for two reasons. Firstly, it captures some realistic aspects of the trade in certain specific markets (e.g., asset markets such as housing and some labor markets). Secondly, it contributes to an understanding of the micro-mechanisms of price formation and their role in shaping market outcomes. In both cases, the features studied are largely neglected in the traditional market equilibrium analyses. The related literature includes the articles by Diamond [4, 5], Diamond and Maskin [6], Mortensen [8, 9], and Zusman and Bell [13]. These articles consider markets of the type described above, in which transactions are concluded at pairwise meetings of agents. The major difference between our work and these articles is in the approach to the basic bargaining problem. In the cited articles it is assumed that a meeting is concluded with an instantaneous agreement which divides the associated surplus in an arbitrary predetermined way (when the surplus is assumed to be divided equally, the division rule is, in fact, Nash's axiomatic bargaining solution). In contrast, the present paper treats the basic bargaining problem with the strategic approach (see Rubinstein [19]) which constitutes an attempt to look into the bargaining black-box. This approach explicitly models the time dimension of the bargaining process, describes in-detail the bargaining procedure, and justifies the agreement as a perfect equilibrium in