In 1988 and 1989, we investigated in three Greek villages the dietary patterns of 182 men and women aged > 70 y by using a validated semiquantitative food-frequency questionnaire and compared these dietary patterns with the traditional Greek diet as ascertained in the late 1950s by Keys and his colleagues. As in the traditional diet, olive oil dominated fat intake, total fat exceeded 35% of total energy intake, average daily consumption of fruits and vegetables exceeded 500 g, and average ethanol intake for men corresponded to two to three glasses of wine per day. In contrast, consumption of meat and meat products has substantially increased and intake of bread and other cereals has apparently declined. We observed no differences between the two periods with respect to consumption of legumes, eggs and egg products, and sugar confectionery. We have also attempted to assess whether a gradient of adherence to the traditional Greek diet can be identified in the diets of the study subjects, and whether it can subsequently predict total mortality. During a follow-up period of approximately 5 y, 53 deaths were observed. The risk of death was apparently higher among the minority of study subjects whose diet deviated substantially from the traditional Greek pattern, compared with the majority whose diet adhered closely to the traditional pattern.