This work undertakes the first comprehensive theoretical assessment of syndicated loans. It is shown that syndicated and bilateral (single lender) loans should be good substitutes in meeting a borrower's financing requirements, but that syndicated loans are more complex and impose additional risks to the parties in the way they are arranged. The existing explantions of loan syndication - that they are hybrids of private bank loans and public debt instruments, that syndication is a portfolio management tool, and that loans are syndicated where they are too large to be provided bilaterally - are unable to substantially explain both the nature of syndicated loans and practice in the loan markets. A rigorous new explanation is developed, which shows that syndication reduces the rate of lending costs, so that the return to the loan originator is greater, and the borrower's cost of financing is lower, where a loan is syndicated rather than provided bilaterally. This explanation is shown to hold in competitive loan markets and to be consistent with the observation that syndicated loans are generally larger than other loans. Incidental to this new explanation, new expressions of the return to a bank from providing a loan on a bilateral basis and from originating a syndicated loan are also developed. New algorithms are also developed for determining the distribution of the commitments from syndicate participants and thus the originator's final hold, the amount it must lend itself, where the loan is underwritten. This provides, for the first time, a rigorous basis for assessing the expected return, and the risk, for the originator of a given syndicated loan. Finally, empirical testing finds that a bank's observed lending history is significant to its decision to participate in a new syndicated loan but that predictions of participation, which are fundamental inputs into the final hold algorithms, based on this information have relatively little power. It follows that there is competitive advantage to loan originators that have access to other, private information on potential participants' lending intentions.