Recent data have shown that individuals with low insulin sensitivity (S(I)) also have reduced whole body maximal oxygen uptake. The objectives of this study were to determine 1) whether muscle mitochondrial function was independently related to S(I) after being adjusted for known determinants of S(I) and 2) whether lower S(I) among African-American (AA) vs. Caucasian-American (CA) women was due to lower muscle mitochondrial function among AA women. Subjects were 37 CA and 22 AA premenopausal women (age: 33.6 +/- 6.3 yr). Mitochondrial function [time constant of ADP (ADP(tc))] was assessed during a 90-s unilateral isometric contraction using (31)P magnetic resonance spectroscopy, S(I) with an intravenous glucose tolerance test, body composition by dual-energy X-ray absorptiometry, and visceral adipose tissue (VAT) with computed tomography. ANOVA was used to compare AA and CA groups, and multiple linear regression modeling was used to identify independent predictors of S(I). Between-race comparisons indicated that muscle oxidative capacity was lower among AAs vs. CAs (ADP(tc): 25.6 +/- 9.8 vs. 21.4 +/- 9.9 s). Multiple linear regression models for the dependent variable S(I) contained 1) VAT and race and 2) VAT, race, and ADP(tc). Significant independent effects for all predictor variables were observed in both the first (r(2) = 0.345) and second (r(2) = 0.410) models. The partial correlation for race was lower in the second model (-0.404 vs. -0.300), suggesting that muscle mitochondrial function contributed to the racial difference in S(I). Lower muscle mitochondrial function among AAs may in part explain lower S(I) among them.