This paper deals with the experimental measurements, data filtering and theoretical representation of the angular position of a human leg in 3-D space during normal and pathological walking. The angular position of a human leg during walking in sagittal plane was measured by a new electrogoniometer made by a UK company named Penny & Giles. This system is a spatial mechanism made of a group oflinks which are coupled by proper angular sensor. This instrument enables an indirect evaluation of the angular position of a human leg in the 3-D space from knowledge ofthe system geometry and from the angular value readings. This instrument is light, small-sized technologically new and is easy to use. However, its dynamics features have not been analyzed in the literature.Therefore we decided to analyze the instrument in order to built a DWT (iscrete avelets Iransform) filter for filtering data recorded by a electrogoniometer Penny & Giles. We built filter corresponding to Daubechies wavelets, DAUB # 20. The DWT filter is sufficient for filtering high frequency noise which exists during experimental measurement ofthe angular position ofa human leg during normal and pathological gait.Filtering using Daubechies wavelets -DAUB # 20 is more efficient than commercial numerical filtering delivered by Penny & Giles company.