A methodology to compare radar reflectivity data observed from two different ground-based radars is proposed. This methodology is motivated primarily by the need to explain relative differences in radar-rainfall products and to establish sound merging procedures of multi-radar observing networks. The authors compare radar reflectivity for well-matched radar sampling volumes viewing common meteorological targets. While spatial and temporal interpolation is not performed in order to prevent any distortion arising from the averaging scheme, the authors considered temporal separation and threedimensional matching of two different sampling volumes based on the original polar coordinates of radar observation. Since the proposed method assumes radar beam propagation under the standard atmospheric condition, we do not consider anomalous propagation cases. The reflectivity comparison results show some systematic differences year to year, but the variability of those differences is fairly large due to the sensitive nature of radar reflectivity measurement. The authors performed statistical tests to check reflectivity difference consistency for consecutive periods.