It has been hypothesized that abdominal obesity leads to insulin resistance partly through decreased adiponectin. However, the cross-sectional and longitudinal associations among waist, adiponectin, and insulin sensitivity have not been examined in older adolescents. Non-Hispanic white and black children were recruited from the Minneapolis school district and underwent three examinations at mean ages 13, 15, and 19. Insulin sensitivity (measured using the gold-standard euglycemic clamp) and waist circumference were measured at all exams. Adiponectin was measured at mean ages 15 and 19. Partial correlations were used to examine associations among waist, adiponectin, and insulin sensitivity at mean age 15 (n = 308) and mean age 19 (n = 218). Longitudinal correlations and a longitudinal regression model were used to predict adiponectin and insulin sensitivity measured at ages 15 and 19, from age 13 waist and change in waist. At age 15, waist and adiponectin were significantly correlated (r = -0.32). At age 19, waist and adiponectin were significantly correlated (r = -0.36), as were waist and insulin sensitivity (r = -0.16). Both baseline waist and change in waist were significantly inversely associated with age 19 adiponectin but with age 19 insulin sensitivity only in men. In conclusion, in adolescents, the association between waist and adiponectin appears to develop several years before the association between waist and insulin sensitivity and there is a longitudinal association between waist and adiponectin. These results support the hypothesis that adiponectin may contribute to the association of waist and insulin sensitivity.