Scholars since Henry Thornton and Walter Bagehot have advised central banks to lend in a crisis to any deserving customer who is illiquid but solvent. Aid should be given quickly, visibly, in large amounts when necessary, temporarily, and comprehensively. The loans should not subsidize errors of judgement, disrupt long-term monetary policy goals, or promote moral hazard. They should restore confidence, prevent panic, and contain spillover. Today, students of the lender of last resort (LLR) also advise that recipients should be chosen equitably and aid be dispensed fairly (see my study with Elizabeth Plautz, 1988). The Federal Reserve, in executing its LLR responsibilities, accepts most of these precepts. Two departures were particularly evident in October 1987, however. First, the Fed prefers to lend to commercial banks and does not lend directly to just any type of organization. Second, it likes to operate behind closed doors. The possibility of further financial deregulation, particularly the union of commercial and investment banking, and the postcrash reform of the stock, options, and futures markets, raise the question whether the Federal Reserve's preferences can continue to be met.