Many recent papers suggest that the basic flex-price target zone model does not perform well empirically. This paper investigates the testable implications of a sticky-price target zone model and finds that, to a limited extent, it has a better empirical performance than the simplest flex-price model. However, in terms of nominal variables, the sticky-price model is found to be observationally equivalent to the flex-price model when the latter is extended to include intramarginal intervention and realignments. Copyright 1994 by Royal Economic Society.