Opposing the life-cycle permanent income hypothesis, empirical evidence suggests large sensitivity of consumption to windfall changes in income. Since the fraction of liquidity constrained households with net worth close to zero is too small to reconcile this observation, Kaplan, Violante and Weidner (2014) argue that collective research has overlooked so called wealthy hand-to-mouth households (W-HtM) by not differentiating between the liquidity of household assets. Holding only little liquid wealth despite owning substantial illiquid assets, W-HtM households are liquidity constrained and sensitive to transitory income shocks. With the objective to provide empirical support, I revisit the W-HtM insight in a context other than the U.S. and test the author’s main results on its reliability by employing their alternative identification strategy of HtM consumption behaviour on longitudinal Italian household data. It shows that HtM households are a stable feature of the Italian population. Pooled over 2002–2016, every fifth household is classified as HtM with 9.3% being poor and 12.3% W-HtM households, whereas the traditional measure based on net worth only identifies a total of 9.6% HtMs. I confirm the observation that W-HtM are similar to unrestricted N-HtM households in terms of demographic characteristics and portfolio composition. Furthermore, this thesis extends the literature by estimating the marginal propensity to consume out of a transitory income shock for the three different household types. As predicted by theory, poor and wealthy HtM households exhibit higher sensitivity compared to N-HtM households. The results are robust for different subsets and modifications of the identification strategy as well as to an analysis of direct survey questions on household’s liquidity.