New Council of Europe regulations mandate housing of two rabbits in the same cage space currently used to house one, provided the animals are socially compatible. This study was designed to assess changes in growth and selected serum chemistry parameters due to pair housing or single housing of rabbits. Six sets of four female siblings of Crl:KBL(NZW)BR rabbits were used. The animals were seven weeks old on arrival. Two siblings of each set were allocated to pair housing, two to single housing. The animals were housed in stainless steel cages (120 cm x 60 cm x 60 cm) with a perforated floor, including a shelf (60 cm x 30 cm) at 30 cm height from the floor. The rabbits were provided with an aspen cube (5 cm x 5 cm x 5 cm), one item per animal. The rabbits were weighed and blood samples were taken from the auricular central artery at four different times during the study. Blood sera were assayed for a set of routinely assayed clinical chemistry parameters: alanine aminotransferase (ALT), aspartate aminotransferase (AST), alkaline phosphatase (APHOS), blood urea nitrogen (BUN), cholesterol (CHOL) and protein (PROT). Mean and variance profiles over the study period were statistically analysed by multivariate analysis of variance. No differences in mean profiles were detected; however, weight (P = 0.0002) and APHOS (P = 0.017) variances were significantly lower in pair-housed animals. The reduction in variance on growth and APHOS attributable to pair housing appears to be rather large. During the 21-week study, occasional fighting was seen between the pair-housed rabbits. After sexual maturity, further major fighting bouts resulted in significant trauma that necessitated the cessation of the study. In conclusion, pair housing appears to have a decreasing effect on growth and APHOS variance, but antisocial behaviour such as fighting remains a serious problem.