This paper presents a novel technique to monitor continuously the azimuthal pointing accuracy of a weather radar antenna. The technique consists of cross‐correlating between modelled and measured echoes from ground clutter in real‐time at low elevation angles under precipitation and non‐precipitation conditions. The azimuthal angle lag with the maximum cross‐correlation indicates the adjustment needed in antenna pointing. The modelled ground clutter echoes were obtained using high‐resolution digital elevation model (DEM) data whereas the measured ground clutter echoes can be obtained in real‐time using a Bayes classifier, which identifies the clutter echoes in the presence of precipitation. The technique has been successfully tested in the Thurnham radar in Southeast England. This method can be used by data users as well as radar operators. It should complement the traditional methods based on sun measurements. Copyright © 2008 Royal Meteorological Society