The total oxygen required for gait (gross utilization) is a widely used indicator of locomotor efficiency. The standard scheme of mass normalization (dividing gross utilization by mass) is intended to eliminate confounding effects that arise from variations in age and size. Despite its prevalence, mass normalization has been shown to be inadequate, as the normalized quantity retains a marked dependence on clinically relevant factors. The current study proposes a new scheme for normalizing oxygen utilization data. The new scheme, called net-nondimensional normalization, is based on the use of net oxygen utilization (gross-rest) and nondimensional gait variables. Measures of statistical significance are employed to demonstrate that net nondimensional normalization is superior to mass normalization. Net nondimensional oxygen utilization is shown to be largely independent of the relevant physiological and anatomical factors, and is therefore well suited for studies where an independent measure of gait efficiency is needed.