Introduction - In gait analysis, numerical representation of graphic waveforms is useful for comparing data between subjects or between groups of subjects. Karhunen-Loeve expansion is an optimal way of representing waveforms concisely and was used previously to represent phasic muscle activity. The expansion is such that only a subset of basis vectors and weighting coefficients are needed to accurately reconstruct the original waveforms which results in a reduction in dimensionality. In this study, we extend the analysis to kinematic and kinetic data acquired during gait analysis.