Pacific Islands populations can be broadly divided into Austronesians (AN) and Non-Austronesians (NAN); obesity and type 2 diabetes are prevalent in the former, although leptin levels in both groups have seldom been investigated. Thirty-seven (20 male and 17 female) adult pairs, matched by age and percent body fat, from AN-speaking Balopa and NAN-speaking Huli, all of whom migrated to settle in Port Moresby, the capital of Papua New Guinea, were selected for comparison of their serum leptin concentrations. The Balopa did not differ significantly from the Huli in age (30.5 +/- 9.7 and 30.0 +/- 8.7 years for males, 33.7 +/- 8.9 and 34.1 +/- 7.5 years for females, respectively) or percent body fat (19.4 +/- 5.6 and 18.8 +/- 4.6 for males, 34.1 +/- 6.2 and 33.3 +/- 5.0 for females), although the BMI of females was lower in the Balopa (26.4 +/- 4.9) than in the Huli (29.7 +/- 4.7) (P = 0.02). In both ethnic groups, females had markedly higher leptin concentrations than males, but there was no significant inter-group difference in males (3.5 +/- 2.6 and 3.1 +/- 4.7 ng/ml, P = 0.14) or females (22.7 +/- 12.9 and 19.7 +/- 11.9 ng/ml, P = 0.40), after controlling for lifestyle factors and serum lipids. Multiple regression analysis revealed that significant predictors of leptin concentration were % body fat (beta = 0.58), sex (male, 0; female, 1; beta = 0.27), and smoker status (non-smoker, 2 0; smoker, 1; beta = -0.15) (R-2 = 0.80), implying that the leptin concentration was primarily determined by lifestyle-derived body fatness. In conclusion, the NAN populations do not endogenously differ in leptin status from the AN populations, who have been recognized as a typical group with a "thrifty" genotype.