The metabolic consequences of the consumption of a diet depend, in part, on the frequency and distribution of meals. The aim of the present study was to examine the number of meals taken per day by elderly persons, and to analyse the relationship between the number of meals per day on the intake of energy and nutrients, the incidence of obesity/overweight, and different cardiovascular risk factors. Food intake was monitored for five consecutive days using 'precise individual weighing' for institutionalised subjects (n = 58), and by means of a 'food intake record' for independent subjects (n = 92). The number of meals taken per day was recorded, along with anthropometric data, blood pressure and serum lipid and lipoprotein levels. No subject took only one meal per day, 7.3% took two, 56.7% took three, 33.3% took four and 2.7% took five. Only 10% of subjects took a mid-morning meal. Breakfast and merienda (a light evening snack) were the meals most frequently omitted. A positive, significant relationship was found to exist between the number of meals taken per day and the intake of fibre (r = 0.2737), thiamin (r = 0.1671), pyridoxine (r = 0.2060), magnesium (r = 0.2423) and the percentage of energy provided by carbohydrates (r = 0.2144). Women subjects who took 2-3 meals per day showed greater body mass indices and a greater proportion were overweight/obese compared to those who took 4-5 meals per day. Subjects who took 4-5 meals per day showed higher levels of HDL-cholesterol and lower levels of LDL-cholesterol/HDL-cholesterol and cholesterol/HDL-cholesterol than did those who took fewer meals per day. Inverse, significant relationships were also found between the number of meals per day and serum cholesterol (r = 0.2297) and LDL-cholesterol levels (r = 0.1984). Taking into account energy and nutrient intakes, the incidence of obesity/overweight and serum lipid and lipoprotein levels, the consumption of 4-5 meals per day would seem more advisable for the elderly than the taking of fewer meals.