Theoretical models of exchange rate determination have not proven useful as accurate predictors of movements in exchange rates. However, these conclusions are the result of evaluation on the basis of statistical criteria of performance; implicitly assuming such criteria are optimal in the subsequent economic decision framework. This assumption is suspect, though, because statistical criteria do not account for market timing is applied in this study to four forecasting models of both the Canadian dollar/US dollar and the West German mark/US dollar exchange rates over the period 1976:11-1984:8. Of these models, only the unconstrained static and dynamic forecasting models exhibited significant markete timing value with respect to the Canadian dollar and then only over a one month time horizon. The constrained dynamic model had market timing abillity at the three month time horizon under Decision Rule I. With respect to the West German mark, the constrained static forecasting model showed market timing ability at both the three and six-month forecast horizon under both decision rules. The static unconstrained model had market timing ability in forecasting the West German mark over a three-month horizon under Decision Rule I.