Objective: We analyzed the relation of osteoprotegerin (OPG) with insulin sensitivity, lipid profile, serum glutamic pyruvic transaminase (SGPT), adipocytokines, and C-reactive protein (CRP) in obese and non-obese subjects. Methods: In the study, 170 subjects (106 obese and 64 non-obese, sex ratio female/maleZ2.03) were included. Thirty-two obese subjects were reevaluated 6 months after the weight loss induced by bariatric surgery. Results: OPG did not differ between obese and non-obese subjects (respective mean values 5.17 and 4.96 pmol/l) or according to gender, but was positively correlated with age (P!0.0001 for both groups). OPG was statistically higher in 18 obese diabetic subjects compared with non-diabetics (PZ0.03). After adjustment for age, no significant correlation was found between OPG and body mass index (BMI), waist, systolic and diastolic blood pressure, cholesterol, triglycerides, high density lipoprotein (HDL) cholesterol, leptin, and adiponectin in both the obese and non-obese subjects. However, OPG was positively correlated with homeostasis model assessment (HOMA) index and SGPT levels in obese subjects at baseline (rZ0.295, rZ0.20, P!0.05) and after adjustment for age (rZ 0.28, rZ0.20, P!0.05). OPG was also significantly correlated with CRP; this correlation persisted after adjustment for age in obese subjects (rZ0.30, P!0.01). In a multivariate analysis in the obese group, HOMA index and CRP were independent predictors of OPG while SGPT was not. Six months post-surgery, OPG did not change, despite a significant reduction in glucose, SGPT, cholesterol, triglycerides, CRP, and leptin values (PZ0.02, PZ0.006, PZ0.007, P!0.001, P!0.001, P!0.001 respectively) and a significant increase in adiponectin and HDL values (P!0.001 for both variables). Conclusion: Our results show that in obese subjects, OPG is not related to BMI. However, we describe new relationships between OPG and both HOMA index and CRP.