This research is concerned with identifying the differing responses of union and nonunion wages to shocks to real output growth, inflation, and the stance of monetary policy. Aggregate measures of union and nonunion wages and salaries are used to construct a time series of the wage differential for several major industrial sectors over the 1976-2001 period. The literature documents the existence of a union wage premium; however, previously the focus has primarily been at the micro-level, and on whether or not a union worker receives greater compensation than an otherwise comparable nonunion worker [e.g., Wunnava and Ewing (1999, 2000)]. Research also links the wage differential to the stage of the business cycle [Wunnava and Okunade 1996] and to the industrial sector [Okunade, Wunnava, and Robinson (1992)]. Theoretical macroeconomic models imply that wages will respond in certain ways to unanticipated changes in aggregate measures of economic activity [e.g., Romer (1996)]. Given the differences in compensation level of union and nonunion workers, and the link to the stage of the business cycle and industry, it is expected that the aggregate wage differentials both for the entire private sector and by industry will respond to macroeconomic shocks in a predictable manner. The relationship among these wage differentials and the macroeconomy is examined in the context of a vector autoregression. In addition, the paper employs the newly developed technique of generalized impulse response analysis [Koop, et al. (1996), Pesaran and Shin (1998)], a method that does not impose a priori restrictions on the relative importance that each of the macroeconomic variables may play in the transmission process. The results show the extent and the magnitude of the relationship between the union-nonunion wage differentials and several key macroeconomic factors. Finally, the paper documents how the responses of these wage differentials vary by industrial sector.