This paper reviews theories that identify motives for mergers, reviews recent empirical research, specifies a model that incorporates alternative motives, and tests the model with data from food manufacturing mergers between 1979 and 1986. Results suggest that capital markets are not efficient, and that mergers are not to redress agency problems. Acquirers paid higher premiums for target firms that have recently had low profitability, and paid higher premiums when the stock market was low. The model explains at best 30 percent of the variation in premiums, suggesting that major explanations for mergers remain as yet unidentified.