The aim of the present work was to evaluate the repeatability and the validity of a short food frequency questionnaire (FFQ) that could be used for older people living in Mediterranean areas. The semi-quantitative FFQ included questions regarding the frequency of consumption of the main food groups and beverages typically consumed in the Mediterranean areas as well as some questions regarding eating habits of older persons. During 2006-2007, for the repeatability assessment (within 10-30 days), 150 individuals (51 +/- 17 yrs, 40% males) were studied, while another 190 individuals (74 +/- 9 yrs, 52% males) were enrolled for the validation purposes. Agreement of the FFQ with the 3-day food records was evaluated using the Bland-Altman method and the Kendall's tau-b coefficient. Repeatability was tested using the Cohen's kappa coefficient. Between 3-day food records and the FFQ, good agreement for alcohol (tau-b = 0.64, p < 0.001) was found, while moderate agreement for food and beverage groups of greens (tau-b = 0.32, p < 0.001), fruits (tau-b = 0.35, p < 0.001), cereals (tau-b = 0.61, p < 0.001), sweets (tau-b = 0.51, p < 0.001), and coffee (tau-b = 0.58, p < 0.001) was observed. Low, but still significant, agreement for fish (tau-b = 0.21, p = 0.001), legumes (tau-b = 0.23, p < 0.001), vegetables (tau-b = 0.23, p < 0.001), pasta (tau-b = 0.25, p < 0.001), potatoes (tau-b = 0.17, p = 0.006) and meat consumption (tau-b = 0.14, p < 0.001) were also found. The FFQ was also valid regarding the estimation of macronutrients and energy intake. Sensitivity analyses by sex, age category (<or= or > 75 yrs), and education status showed similar validity of the FFQ in each subgroup, except for elders older than 75 years. The repeatability of the FFQ was fair in all foods tested (Cohen's kappa coefficients varied between 0.15-0.39, p-values < 0.05). The suggested FFQ seems to be a reasonably valid and repeatable measure of dietary intake and can be used in older persons living in the Mediterranean areas.