LOFTIN, MARK, MELINDA SOTHERN, LAURA TROSCLAIR, ANN O'HANLON, JAMES MILLER, AND JOHN UDALL. Scaling VO 2 peak in obese and non-obese girls. Obes Res. 2001;9:290 -296. Objective: The conventional ratio method (milliliters O 2 per mass) typically is used to express VO 2 peak. The goal of the current study was to compare VO 2 peak of obese girls with normal-weight girls by ratio and allometric scaling methods. Research Methods and Procedures: We compared VO 2 peak by ratio and allometric methods in 46 obese and 47 normal-weight girls. Indirect calorimetry was used to measure VO 2 peak during either treadmill running or walking. Regression analysis was used to determine coefficients for mass and stature for each group with ANOVA used to compare data between groups. Results: The obese girls were taller and had higher values of body fatness (p Յ 0.05). Absolute VO 2 peak (liters per minute) was similar between groups; however VO 2 peak relative to mass was 50% lower (p Յ 0.05) in the obese girls. When VO 2 peak (milliliters per minute per kilogram) and mass were correlated, r ϭ Ϫ0.48 was found in the obese group. Allometric scaling of logarithmic transformed stature and mass reduced this to r ϭ Ϫ0.002, thus eliminating the bias associated with the ratio method. Adjusting VO 2 peak allometrically scaled for mass, stature, and the combination of mass and stature reduced the difference between groups from 50% (ratio method) to 10% to 11% (p Յ 0.05) with higher values found in the normal-weight girls. Discussion: These results demonstrate the bias associated with the ratio method when comparing VO 2 peak in obese girls with VO 2 peak in normal-weight girls. Allometric scaling eliminated the bias and thus may reflect a truer comparative response.