Mean platelet volume (MPV) is increased in patients with coronary heart disease or at risk for stroke. However, MPV determinants have never been assessed in a population study. The present investigation is a cross-sectional study involving 366 non-selected subjects (both sexes, mean age 72.9 ± 5.5 [1 SD] years). The main cardiovascular risk factors, several indexes of adiposity (including percent body fat as estimated by skinfold measurement, and ultrasound detection of hepatic steatosis and thickness of abdominal subcutaneous and visceral fat) and ischaemic electrocardiographic (ECG) changes were assessed in all subjects. Platelet parameters were determined by a Bayer ADVIA 120 counter. In addition to being associated directly with platelet distribution width (PDW) and inversely with platelet count (p<0.0001 for both), MPV values were associated with subcutaneous abdominal fat (p=0.02), fasting blood glucose (p=0.002) and the prevalence of ischaemic ECG changes (p=0.004), and tended to be higher in the subjects with a greater prevalence of hepatic steatosis (p=0.07) and higher Homeostasis Model Assessment (HOMA) index (p=0.09). In multiple logistic regression, of the non-platelet parameters only percent body fat (p=0.006), ischaemic ECG changes (p=0.01) and blood glucose (p=0.03) remained independently associated with an MPV ≥8.4 fl (high tertile). The relative risk (odds ratio) of having ischaemic ECG changes for the subjects with MPV ≥8.4 fl was 4.2 (95% confidence interval: 2.5–7.1; p=0.006) with respect to the subjects with lower MPV values. Blood glucose, percent body fat and ischaemic ECG changes were the main MPV determinants in our elderly population.