Data from a crossbreeding experiment conducted during 1992–97 involving 31 genotypes from tropically adapted British (B), Sanga-derived (S), Zebu cross (Zx), Zebu (Z), and Continental (C) beef cattle breed groups were analysed to compare least-squares means, direct and maternal genetic effects, and heterosis estimates for birthweight, weaning weight, yearling weight, final weight (18 months), and pre- and post-weaning average daily gain (ADG). The genotypes were regrouped as Bos taurus (B, S, C)- and Bos indicus (Z, Zx)-derived groups to enable the comparison of direct (dD) and maternal (mD) dominance effects among indicine (II), taurine–indicine (TI), and taurine (TT) crosses. Genotype, contemporary group (year of birth, season of birth, and age of the dam), sex, and genotype × sex interactions were significant (P < 0.01) sources of variation for all the traits. Treatment to control parasites significantly (P < 0.01) affected post-weaning growth traits. In general, crossbred calves performed better than purebred calves. Z dam breeds resulted in lower birthweight, and Z sire breeds and S dam breeds resulted in heavier birthweights. For traits after birth, ZC and ZC crosses with S and Z showed heavier weights and higher gains. Prior to weaning, males weighed significantly more and gained weight at a faster rate than females in most of the crossbreds. Weight gain was relatively low between weaning and yearling age.Direct and maternal additive effects were estimated as a deviation from the British breed group mean for various traits. Direct additive effects of C, Z, and S were high and significantly different from the British mean for all the growth traits. Maternal additive effects of C were low and not significantly different from the British mean. Large negative maternal additive effects of Z and Zx caused lower birthweights of calves from Z and Zx dams. A decrease of maternal additive effect from weaning to final weight and pre-weaning to post-weaning ADG was noticed. The magnitude of dD effects was higher in TI crosses than in II crosses for all the traits except for birthweight, indicating the advantage of Bos taurus × Bos indicus crosses. In TT crosses, dD was only significant for weaning weight (P�<�0.01) and pre-weaning ADG (P < 0.05). Significant (P < 0.01) and positive mD effects observed in TI crosses indicated a better maternal environment provided by crossbred dams. High correlation coefficient estimates (0.92–0.99) between least-squares means and predicted means, observed for a set of F1 genotypes, indicated the prediction of performance of untested genotypes with reasonable accuracy. The per cent heterosis estimates were higher in Zebu × British breed crosses.