This study examines the relationship between the probability and extent of a firm's participation in interestrate swap markets and the magnitude of some of its operating and financial ratios. These ratios are selected to proxy incentives to reduce agency costs of leverage and information asymmetry in credit markets. The data on swap usage are obtained from disclosures mandated by SFAS 105. The results suggest that fixed-rate payers, compared to non-swap-users, have more leverage, greater profitability, more growth options, less operating risk, lower ratios of fixed to total assets, and more divergent earnings estimates. On the other hand, floating-rate payers do not seem to have financial and operating characteristics significantly different from non-swap-users, although they do have less divergent earnings estimates.One of the most important developments in the financial services industry in the last decade is the spectacular growth in the market for interest-rate swaps. The volume of outstanding swaps in 1992 (measured in terms of notional principal) is an estimated $2.3 trillion. 1 Conventional explanations of the growing popularity of interest-rate swaps are often based on their supposed ability to reduce effective costs of borrowing through financial arbitrage. These "explanations" of the economic value of swaps usually rely on the existence of an ad hoc market imperfection such as a quality spread differential between the terms of loans contracted at fixed and floating interest rates, which results in a comparative advantage for a firm in one market. 2 However, Turnbull (1988) and Rendleman (1992) demonstrate clearly that, in the absence of externalities, an interestrate swap is a zero-sum game between counterparties, and conventional explanations of the tangible benefits of swaps are therefore not convincing. More satisfactory explanations have relied on the role of swaps in mitigating information asymmetry in credit markets (Arak et al, 1988;Titman, 1992), reducing agency costs of leverage (Wall, 1989), expropriating guarantor wealth (Rendleman, 1992), and the special treatment of swaps in bankruptcy proceedings (Litzenberger, 1992). This study investigates whether there is evidence in cross-sectional data to support the agency cost and information asymmetry explanations for swap usage.