Using an innovative dataset built by merging survey and administrative data, we provide new estimates of intergenerational earnings' inequality between fathers and sons in Italy, extending previous evidence in several directions. We rely on the TSTSLS method to predict fathers' earnings and compute intergenerational elasticities and imputed rank-rank slopes, trying to reduce estimation biases. Confirming previous evidence, we find that Italy is characterized by a high intergenerational inequality in crosscountry comparison. Extending previous analyses, we show that the intergenerational association increases when sons at older ages and multi-annual averages of pseudo-fathers' and sons' earnings are considered. We also find that the intergenerational persistence differs across geographical macro-areas and is high also for daughters, especially when family earnings are considered. Furthermore, estimates where possible mediating factors of the parental influence are included among the covariates show that a high intergenerational association persists when sons' education and occupation are controlled for.JEL Codes: J24, J31, J62