Health in relation to tail-biting behaviour was investigated on a problem farm. Quartets (n = 16) of age- and gender-matched fattening pigs including a tail biter (TB, n = 16), a victim (V, n = 16), a control in the same pen (Ctb, n = 10) and a control in a pen where no tail biting was observed (Cno, n = 14) were chosen by direct behavioural observation. Haematological and clinicochemical analyses, autopsy and histological examination of 16 different tissues were carried out. Tail lesion severity was evaluated both macroscopically, on the basis of inspection, and histologically, in the sagittally cut tail. Category effects were tested using Friedman's ANOVA by Ranks, Cochran's Q or a repeated-measure GLM and, if significant, pair-wise tests were conducted using Wilcoxon Signed Ranks or McNemar's Test. The number of received tail bites correlated better with histological than with macroscopic tail lesion scoring because of deep inflammation beneath healthy skin in some cases. Most individuals had mild inflammatory lesions in internal organs suggestive of generalized activation of the immune system, and 30% of the animals were anaemic, possibly because of systemic spread of infectious agents. V had more severe respiratory organ lesions and higher serum protein concentrations than all other categories of pigs. Liver- and muscle-specific enzymes (alanine aminotransferase, alkaline phosphatase and creatine kinase) differed between categories. In conclusion, most animals had signs of generalized activation of the immune system, possibly because of systemic spread of infectious agents. V pigs suffered from more severe inflammatory lesions than TB, Ctb or Cno. Deep infections may exist under healthy skin in the tail of bitten pigs.