Topics in antitrust theory, empirical analysis, and policy are examined, with an application to the USAir-Piedmont merger. The first chapter presents Cournot models to assess the welfare tradeoff between increases in market power and increases in efficiency that may result from a merger. The magnitudes of efficiency gain sufficient to satisfy two alternative welfare standards are derived. To satisfy the first standard, the efficiency gain must be sufficient to offset the deadweight loss created by increased market power. To satisfy the second standard, the efficiency gain must be sufficient to offset the incentive to increase prices resulting from increased market power. In both cases, the critical efficiency gains derived offset the effects of the mergers that are determined endogenously within the model, rather than effects that are predetermined. Results are presented for a range of assumptions about the number of firms, the pre-merger markup· of price over marginal cost, and fixed costs. The second chapter evaluates the method of residual demand elasticity analysis. The relationship between residual demand elasticities and market power is examined. Alternative methods for making inferences of post-merger market power using estimates of pre-merger residual demand elasticities also are examined. The third chapter estimates and analyzes residual demand elasticities for the USAir-Piedmont merger. Five routes on which the two firms competed prior to the merger are examined. Pre-merger time-series data is used to estimate pre-merger, post-merger, and market residual demand elasticities for each route. The results suggest that a significant degree of market power existed on some routes prior to the merger and that increased market power appears likely to result from the merger on some routes. Other routes exhibit highly elastic residual demand in spite of high levels of market concentration.