This paper develops and tests an analytical framework for analyzing the selection of tactics in bargaining. Using a variant of power-dependence theory, the authors propose that bargainers will use different dimensions of dependence, such as the availability of alternative outcomes from other sources and the value of the outcomes at stake, to select among different tactics. To test this model, the authors conducted two simulation experiments that portrayed an employee-employer conflict over a pay raise, manipulating four dimensions of dependence: employee's outcome alternatives, employee's outcome value, employer's outcome alternatives, and employer's outcome value. Within this context, respondents estimated the likelihood of each actor (employee, employer) adopting four tactics: self-enhancement, coalition, threat to leave, and conflict avoidance. The results of one experiment show that an actor's own dependence, rather than his opponent's dependence on him, is the primary basis for his evaluation and selection of tactics, and also that decisions regarding different tactics are determined by different dimensions of dependence. The results of the other experiment indicate that the opponent's initial lactic affects the links between dimensions of dependence and an actor's tactics, and the dimensions of dependence affect the propensity toward "tactic matching."