The objective is to investigate the relation between the levels of two serum adipocytokines (adiponectin and resistin) and non-alcoholic fatty liver disease (NAFLD) in obese children. In this study, 113 obese children were enrolled and divided into 3 groups. Obese group 1 was defined as obese children without any liver abnormality. Obese group 2 was defined as obese children just with fatty infiltration of the liver in ultrasonic appearance and obese group 3 was defined as obese children with liver function abnormality. The controls consisted of 37 nonobese children without endocrine, metabolic or kidney disease. The levels of serum adiponectin and resistin were measured by ELISA method. Insulin resistance by homeostasis model (HOMA-IR), area under curve of glucose (AUCG), serum total cholesterol, triglyceride, alanine aminotransferase, uric acid, HDL-cholesterol, LDL-cholesterol and body mass index (BMI) were measured as well. In obese children, NAFLD were found in 63 cases (55.75%). Serum adiponectin levels of obese children were significantly lower than that of controls (3.63 vs 5.79 μg/mL, P<0.001) while serum resistin levels were not different (P = 0.876). Moreover, serum adiponectin levels in obese group 1 were significantly higher than that of group 2 and 3 (4.24 vs 3.37 and 3.12 μg/mL, all P<0.05) and no difference was found between obese group 2 and obese group 3 (P>0.05). Serum resistin levels among the three obese groups were 4.37 ng/mL, 3.72 ng/mL and 4.24 ng/mL without significant difference (P = 0.592). NAFLD, BMI, gender and HDL-cholesterol were independent determinants of serum adiponectin levels in children analyzed by multiple regression analysis, which explained 33% of the variance. Serum adiponectin levels were inversely associated with BMI, gender and NAFLD (all P<0.05) and were positively associated with HDL-cholesterol levels (P = 0.033). These results suggest that adiponectin might be a protective factor in NAFLD occurrence in obese children, and that the measurement of adiponectin should be part of the standard evaluation of the obese child and may help to evaluate the occurrence of NAFLD.