The aim of the present study is to assess the quality of the variables "name" and "surname" recorded in the mortality records that the Italian Institute of Statistics (ISTAT) releases to its regional offices since 1999. These records could both constitute a regional mortality database in those regions that do not have one, and be a useful tool for ascertainment of vital status in follow-up studies. The study was conducted in Turin, North West Italy, through record linkage between ISTAT database and the local population register that records all deaths occurred among residents. This was considered as the gold standard. Firstly the concordance of name and surname was studied; this was 92% if the full length of name and surname was used raising to 97% using substrings of the two information. Secondly the cohort of 1999-2001 residents in Turin was linked to the ISTAT database using a four step record linkage with four different keys. 94,3% of subjects were correctly found. Within those not linked there was a higher proportion of subjects born abroad, women, young and unmarried. We conclude that the quality of ISTAT database should be further improved before implementing its use for follow-up studies.