This paper studies the weighting factor of the two-dimensional cepstrum (TDC) distance measure. By using the inverse standard deviation of feature as weighting factor, a vector quantization (VQ) based speech recognition method is used to recognize TDC speech pattern. The advantage of this speech recognition method is its simple complexity in implementation. The effect of the weighting factors on the TDC distance measure is also investigated. It shows that the performance of the inverse standard deviation is about 2% better than the one derived from quefrency weighting and time derivative weighting.