The aim of the present analysis was to evaluate the association of the Mediterranean diet (MeDi), smoking habits and physical activity with all-cause mortality in an Italian population during a 20-year follow-up study. A total of 1693 subjects aged 40–74 who enrolled in the study in 1991–5 were asked about dietary and other lifestyle information at baseline. Adherence to the MeDi was evaluated by the Mediterranean dietary score (MedDietScore). A healthy lifestyle score was computed by assigning 1 point each for a medium or high adherence to the MedDietScore, non-smoking and physical activity. Cox models were used to assess the associations between lifestyle factors and healthy lifestyle scores and all-cause mortality, adjusting for potential confounders. The final sample included 974 subjects with complete data and without chronic disease at baseline. During a median of 17·4 years of follow-up, 193 people died. Subjects with high adherence to the MedDietScore (hazard ratio (HR) 0·62, 95 % CI 0·43, 0·89)), non-smokers (HR 0·71, 95 % CI 0·51, 0·98) and physically active subjects (HR 0·55, 95 % CI 0·36, 0·82) were at low risk of death. Each point increase in the MedDietScore was associated with a significant 5 % reduction of death risk. Subjects with 1, 2 or 3 healthy lifestyle behaviours had a significantly 39, 56, and 73 % reduced risk of death, respectively. A high adherence to MeDi, non-smoking and physical activity were strongly associated with a reduced risk of all-cause mortality in healthy subjects after long-term follow-up. This reduction was even stronger when the healthy lifestyle behaviours were combined.