This work has a dual purpose. First of all, we investigate the effect of uncertainty on economic activity in Brazil. In order to do that, we construct several proxies which seek to capture the uncertainty level prevailing in Brazil (domestic uncertainty) and in several of our major trading partners (external uncertainty). Next, we estimate vector autoregressive (SVAR) models, in line with Baker, Bloom and Davis (2016). The results suggest that uncertainty has, in fact, contractionary effects on the activity in Brazil. Second, we study the effect of uncertainty on effectiveness of monetary policy in Brazil. Thus, we make use of interacted vector autoregressive (IVAR) models, as proposed by Aastveit, Natvik and Sola (2013), estimated, however, by Adaptive LASSO. Our estimates do not corroborate the hypothesis that under high uncertainty the effects of monetary policy on the activity are lower than under low uncertainty.