This study aims at quantifying the genetic basis of tail-biting receipt in the Tai Zumu population. Tail-biting receipt was recorded at the end of the fattening period on 33,266 gilts that were raised in groups of 6 to 20 females among 3 nucleus herds. The phenotype under measure was the presence of lesion(s) at the end of the fattening period, reflecting that the gilt did not defend itself from tail biters attacks. The objectives were to: (1) compare genetic merit for growth and leanness between non-bitten females and bitten females; (2) quantify the contribution of social genetic effects to the phenotypic variation of the trait ‘bitten’; and (3) estimate the effect of the environment on the genetic expression of this trait as interactions between genetics and environment (G×E). The prevalence of tail bites varied from 2.8 to 10.8% among herds that differed according to the groups’ characteristics (e.g. size, number of mixings) and feeding system during fattening. Models for tail-biting receipt included the fixed effects of number of group mates, (herd)-year-month, and the random effects of group, litter and single direct genetic effects or also social genetic effects. The model was applied to the complete gilt population and separately to 2 sub-populations, corresponding to the 2 nucleus herds where tail-biting receipt was highest (>5%). Bitten females had a higher genetic merit for growth and leanness than non-bitten females. Tail-biting receipt was analysed as a linear trait, following a normal distribution with the restricted likelihood methodology. Social genetic effects contributed 81 to 93% of total heritable variance, which equalled 40 to 80% of phenotypic variation, whereas only 6-8% was explained with a direct model. Differences in the ranking of connecting sires between herds according to their direct, social and total breeding values yielded rank correlations not different from zero, indicating strong G×E. Selection against tail biting is possible even though the direct genetic contribution for this trait is low. Considering social genetic effects improves quantification of heritable variation, and accounting for them in breeding schemes would increase response to selection against this deviant behaviour.