The paper questionsthe use of Total Factor Productivity (TFP) measures as a proxyfor the technological level in Italian agriculture. TFP growthreveals a cyclical behaviour and a short-run component due toshocks on both the demand and supply sides. In order to identifyreal long-run technological change, a Kalman filter procedureis applied to a stochastic process depicting the generation oftechnical change induced by R&D and Extension expenditure.The empirical evidence reveals that short-term shocks greatlyaffect the traditional measure of productivity. A better indicatorof technological progress can be obtained by estimating the long-runcomponent of productivity, which seems to be significantly inducedby R&D-Extension public expenditure.