Leptin's association with fasting insulin raises the possibility that hyperleptinaemia is an additional component of the Metabolic Syndrome, or perhaps underlies the syndrome. This population-based study of Western Samoans examined the relationship of serum leptin with insulin sensitivity assessed by Homeostatic Model Assessment (HOMA) and components of the Metabolic Syndrome. Two hundred and forty subjects (114 men, 126 women), aged 28-74 years, were drawn from a study conducted in 1991. An oral glucose tolerance test indicated that 59 subjects had diabetes. Diabetic men had higher leptin levels than non-diabetic (6.0 vs 3.2 ng ml-1) but this difference was no longer significant after adjustment for BMI. Leptin levels in diabetic women (24.7 ng ml-1) non-diabetic women (22.6 ng ml-1) were not different. Leptin was strongly, positively correlated with BMI, fasting insulin and mean blood pressure after adjusting for age and sex (r > 0.43, p < 0.001), irrespective of glucose tolerance status. Linear regression models indicated that leptin was associated with insulin sensitivity independent of age, BMI, waist/hip ratio, triglycerides, HDL-cholesterol, and hypertension. Similar models were computed with mean blood pressure or triglycerides as the dependent variable, and including insulin sensitivity with the independent variables. Leptin was independently associated with mean blood pressure in men, but was not independently associated with triglycerides. Mean levels of 2-h insulin, triglycerides, LDL-cholesterol, and systolic blood pressure varied across tertiles of leptin in men after adjusting for age, BMI, and insulin sensitivity, and mean levels in the top tertile tended to be higher than in the lowest tertile. These results indicate an independent relationship between leptin and insulin sensitivity, but the equivocal results concerning associations of leptin with components of the Metabolic Syndrome make it unlikely that leptin affects these directly.