The Mediterranean diet (MD) has been associated with prolonged survival in the general population, but no meta-analysis has apparently investigated the potential health benefits in relation to mortality in the elderly. We performed a longitudinal analysis on 5200 individuals aged ≥65 years identified within the general population recruited in the Moli-sani study cohort (2005–2010). Adherence to the MD was appraised by the a priori Mediterranean diet score (MDS; range 0–9). Survival estimates were derived using Cox regression and competing risk models. For the meta-analysis, PubMed and Scopus databases were searched from inception until April 2018 to identify prospective studies on the MD and death risk in the elderly. Over a median follow-up of 8·1 years, a total of 900 deaths were ascertained in the elderly sub-sample of the Moli-sani cohort. A one-point increase in the MDS was associated with lower risk of all-cause, coronary artery disease/cerebrovascular and non-cardiovascular/non-cancer mortality (multi-variable hazard ratio (HR)=0·94; 95 % CI 0·90, 0·98; HR=0·91; 95 % CI 0·83, 0·99 and HR=0·89; 95 % CI 0·81, 0·96, respectively). In a meta-analysis of seven prospective studies, including our results, for a total of 11 738 participants and 3874 deaths, one-point increment in MDS was associated with 5 % (4–7 %) lower risk of all-cause death. An inverse linear dose–response relationship was found from a meta-analysis including three studies. In conclusion, a prospective cohort study and a meta-analysis showed that closer adherence to the MD was associated with prolonged survival in elderly individuals, suggesting the appropriateness for older persons to adopt/preserve the MD to maximise their prospects for survival.