It has previously been demonstrated that plasma leptin correlates to body fat content. Increased body fat content is accompanied by low insulin sensitivity, which is compensated with increased insulin secretion. We therefore studied whether plasma levels of leptin also correlate to insulin secretion and sensitivity in humans. Therefore, we examined insulin sensitivity by the euglycemic-hyperinsulinemic clamp technique and measured the insulin response to intravenous arginine (5 g) at fasting and 14 mmol/l glucose in postmenopausal women. Percent body fat content was determined with impedance measurements. Log plasma leptin significantly correlated to percent body fat (r = 0.84, P < 0.001). In women with normal glucose tolerance (n = 36), partial correlation studies controlling for body fat content revealed significant correlations between log plasma leptin and fasting insulin levels (r = 0.39, P = 0.029), the insulin response to arginine at both glucose levels (r = 0.38 and r = 0.37, P < 0.036 for both), and the glucose potentiation of arginine-stimulated insulin secretion (r = 0.40, P = 0.025). In contrast, in women with impaired glucose tolerance (n = 17), these correlations were not significant. Plasma leptin did not correlate with insulin sensitivity independently of body fat content. To study whether the correlation between leptin and insulin would be explained by insulin stimulating leptin secretion, we examined plasma leptin during hyperinsulinemic conditions (689 +/- 41 pmol/l), under both euglycemia (5.0 mmol/l, n = 10) and hypoglycemia (2.5 mmol/l, n = 7). However, under both these conditions, plasma leptin was unaltered. In conclusion, plasma leptin 1) reflects body fat content and 2) correlates to insulin secretion independently of percent body fat in postmenopausal women with normal glucose tolerance.