A major aim of recent empirical modelling of the business cycle isto identify the relative importance of aggregate supply and demandshocks. Supply or technology shocks are associated with permanent(structural) effects on economic activity whereas demand shocks arerelated to temporary (cyclical) effects. Most studies in this veinuse multivariate VAR-models or the common trends-cointegrationapproach in order to disentangle supply and demand shocks. As analternative, this paper uses the methodology of unobserved (orstructural) components time series models as set out in Harvey(1989) for identification of technology and demand shocks in atwo equation system of labour produc-tivity and industrial output.The novelty is the introduction of correlation between the twotypes of shocks such that the mutual dependency of these shockscan be estimated explicitly. This is because tech-nology shockswill have cyclical (temporary) effects, and demand shocks willhave structural (permanent) effects, which are not fully describedby the interaction of the endogenous variables in the model. Theestimation procedure is set out in Koopman et al. (1995). The datais quarterly time series of labour productivity and industrialoutput for Germany, The Netherlands, the United Kingdom and theUnited States. Our results show that the covariance of the dynamicsof supply and demand shocks appears to be important in thesecountries. It indicates a good coordination is needed betweenstructural and cyclical policies.