A calf’s weight at weaning is influenced not only by its own genes for growth (direct genetic effects) but also its mother’s genes for maternal (milking) ability. Selection for genetic improvement in both maternal ability and growth are an important part of current breeding programmes for many beef breeds. Estimated breeding values for both components are obtained by separating the direct and maternal effects on live weights around weaning by fitting appropriate statistical models. Many previous studies have estimated large negative genetic correlations between the direct and maternal components (rAM) of a round -0.5 (Meyer, 1997). Whether these estimates are a true reflection of the actual biological relationships have frequently been questioned. Meyer (1997) showed that by fitting additional random terms in the model [i.e. sire or sire*year environmental terms, dam-offspring environmental covariance/correlation (rAeDe)] the magnitude of the negative rAM estimates could be reduced. The aims of this study were: (i) to estimate the magnitude of the rAM in a UK population, (ii) to investigate the relative importance of various additional terms, and (iii) to test the importance of including a non-zero rAM in these alternative models.