It is clear that at present various versions of the Calvo (1983) model of price adjustment are dominant in monetary policy analysis'see, e.g., Woodford (2003). This is true despite well-known criticisms including Mankiw (2001) or Mankiw and Reis (2002) and the well-documented need for the addition of ad-hoc features if actual inflation and output data are to be matched. Accordingly, there is ample reason, to give consideration to alternative models. In this paper, a new look is given to the P-bar model utilized by McCallum and Nelson (1999a, 1999b), based on previous work by Mussa (1981) and others. Relative to the Calvo model, the P-bar specification has three significant advantages: it satisfies the strict version of the natural rate hypothesis; it relies on costs of adjusting output, which are more tangible than menu costs of changing prices; and its basic version produces more realistic autocorrelation patterns than does the basic Calvo specification. The present paper develops these comparisons more completely and systematically than in previous work.