This paper presents a novel approach to quantitatively analyze the error caused by neighbor decimation in the conversion between arbitrary sampling rates. Based on a fractional sampling rate conversion structure, we propose an arbitrary sampling rate conversion algorithm for software radio by neighbor decimation, realize the structure and deduce its computation complexity. Farther, we quantitatively analyze the error of neighbor decimation in detail. Treated as a kind of SNR reduction, the effects of neighbor decimation can be a guideline of parameter selection. The conclusion proved that this method has superiority both in currency and efficiency, the currency behaved as a uniform structure to realize all arbitrary sampling rate conversion, and the efficiency exhibited by the equivalent computation complexity to traditional methods.