During growth (27 to 75 days of age), a total of 384 rabbits were kept in 72 individual cages, 48 bicellular cages (2 rabbits/cage) and 24 collective cages (9 rabbits/cage). To evaluate the effects of the housing system on the fear level and behavioural patterns of rabbits at the two ages (39 to 45 days and 66 to 73 days), a tonic immobility test and an open-field test were conducted and their behaviour was video recorded. In the tonic immobility test, the number of attempts to induce immobility (1.38) was lower, and the duration of immobility (47.8 s) was higher (0.05 < P < 0.01) in the rabbits housed in individual cages than in those kept in bicellular (1.72 attempts and 25.0 s of immobility) and collective cages (1.99 attempts and 25.0 s of immobility). During the open-field test, the rabbits from individual and bicellular cages showed higher latency (38.8 and 40.3 v. 27.0 s), a lower number of total (73.3 and 81.7 v. 91.9) and central displacements (3.6 and 2.8 v. 5.4) and a shorter running time (11.8 and 13.6 s v. 17.7 s) and the time biting the pen (5.5 and 9.1 s v. 28.2 s) compared with the rabbits kept in collective cages (0.05 < P < 0.001). During the 24-h video recording, the rabbits in individual and bicellular cages spent less time allogrooming (0.34% and 0.19% v. 1.44%), moving (0.74% and 0.60% v. 1.32%) and running (0.08% and 0.03% v. 0.21%) than the rabbits in the collective cages (0.01 < P < 0.001). The lowest numbers of alerts and hops were observed in the rabbits kept in bicellular cages. With increasing age, a lower number of rabbits were sensitive to the immobility test and more rabbits entered the pen spontaneously during the open-field test (P < 0.001). In conclusion, the rabbits in individual cages exhibited the highest fear level and incomplete behavioural patterns; the rabbits housed in collective cages showed the lowest fear levels and had the possibility of expressing a wider range of behaviour; and the rabbits in bicellular cages exhibited an inconsistent pattern of fear in the tonic immobility and open-field tests. Probably, these rabbits were in a less stressful condition compared with animals in individual cages because social contacts were allowed, even if freedom of movement was more limited.