Two speculative strategies within the European Monetary System are empirically evaluated. The potential profitability of speculating on a currency's devaluation at a realignment crucially depends on being able to predict timing and magnitude of the parity change. Such opportunity has been eliminated from the system since 1983. For the reverse strategy of "borrowing low, investing high," the evidence since 1983 suggests significant profitable opportunities for the weaker EMS countries - Belgium, Denmark, France and Italy - unconditional on knowledge of the timing of realignments. We conclude that this is due to a "peso problem" type of premium. © 1993 Kluwer Academic Publishers.