THE ST. PETERSBURG PARADOX AND PASCAL'S WAGER JEFF JORDAN Pascal's Wager purports to give a good reason to believe in God.1 The basic idea of the Wager, put simply, is that if God exists, then the utility of belief is infinite; while if God does not exist, the loss for both the believer and the unbeliever is but finite. Since the expected utility of belief is greater than nonbelief, one has a good reason to try to bring about theistic belief.2 The notion of an infinite utility plays a prominent role in the Wager; without that notion the Wager is clearly unsound. Not surprisingly, the notion of an infinite utility has also been the object of philosophical suspicion. One particularly interesting objection to the Wager's use of the notion of an infinite utility is built upon the St. Petersburg paradox. This paradox was formulated by Nicholas Bernoulli in correspondence with Pierre Montmort in the early 18th century; and was the occasion for the formulation of that famous doctrine of economics, the law of declining marginal value. 3 In broad terms the paradox holds that it is not true that as long as the expected utility of an act A is infinite, reason demands that one do A at any finite cost. From this it's concluded that the notion of an infinite expected utility (and by extension, the notion of an infinite utility) is problematic and best discarded. Given this conclusion, Pascal's famous Wager seems doomed to failure from the start. In what follows I argue that the St. Petersburg paradox can be defused without resorting to the wholesale exclusion of infinite utilities. The paradox is defused via a decision-theoretic principle, the Sure Loss Principle, which answers the paradox in a way which is compatible with infinite utilities. From this it follows that if Pascal,s 207