Few works more than Esping-Andersen's 'Three Worlds of Welfare Capitalism' have drawn researchers' attention on institutional features that characterize the diverse typologies of welfare regimes; yet the impact of the different institutional settings on income distribution has mostly been taken for granted. This paper uses cross nationally comparable data from the Luxembourg Income Study (LIS) to assess the performance of Italian Welfare against that of other countries of the 'conservative' cluster (Germany, France and the Netherlands), as well as countries of the 'liberal' (United Kingdom) and 'social-democratic' (Sweden) regime. The results reveal that the institutional setting is strongly correlated with the antipoverty efficiency of social transfers (indeed in the direction one would expect), with the significant exception of Italy, whose performance approaches that of the United Kingdom. The reason for such performance must be found within the structure of social transfers. Italy's extremely generous pension system has in fact crowded out any form of support directed to younger families, both in child support or means-tested assistance. While two-earners-households find a good hedge against poverty risks in market income, one earner household shaped around the 'single male breadwinner model' face considerable poverty risks, which increase linearly by a factor of 10% according to the number of children in the household. As a consequence of the greater diffusion of such family model in Italy, child poverty rates tend to be double and almost ten times higher than the Swedish ones, approaching the British standard (almost 20%). Evidence also shows that what prevents a greater diffusion of the double income household is probably not as much related to cultural inheritance as to concrete difficulties encountered by women in young households combining parental and working responsibilities. In higher quintiles of disposable income, where services to families may be easily acquired on the market, female employment ...