OBJECTIVE: To prospectively evaluate the association of three dietary patterns: the MIND (Mediterranean-DASH diet intervention for Neurodegenerative Delay) diet; a Mediterranean-type diet and a traditional diet, with all-cause mortality over a 12-year period in an older sample. DESIGN: A longitudinal birth cohort study. We ascertained dietary patterns using FFQ data at baseline (2004-2007) and mortality using linkage data. Cox regression was used to estimate mortality hazard ratios (HR) with adjustment for confounders. SETTING: The Lothian Birth Cohort 1936 (LBC1936) study in Edinburgh, Scotland. PARTICIPANTS: Dietary patterns were ascertained in 882 participants, mean age 69·5 (±0·8) years, at baseline. During the 12-year follow-up (to October 2019), 206 deaths occurred. RESULTS: In the basic-adjusted model, all three dietary patterns were significantly associated with mortality, the MIND diet and Mediterranean-type diet with a lower risk and the traditional diet with a higher risk. In fully adjusted models, MIND diet score was inversely related to all-cause mortality (HR 0·88; 95 % CI 0·79, 0·97) such that the risk of death was reduced by 12 % per unit increase in MIND diet score. Participants in the top compared with the bottom third of MIND diet score had a 37 % lower risk of death (HR 0·63; 95 % CI 0·41, 0·96). No significant associations with the Mediterranean-type or traditional dietary patterns were observed in the final multivariate model. CONCLUSIONS: Our findings suggest that closer adherence to the MIND diet is associated with a significantly lower risk of all-cause mortality, over 12 years of follow-up, and may constitute a valid public health recommendation for prolonged survival.