To determine the effect of age at castration on physiological and immunological stress indices, 60 Holstein-Friesian bull calves were sourced so that they were in one of five age groups for Burdizzo castration on day 0 (16 July 2002; no. = 10 per treatment) : 1·5, 2·5, 3·5, 4·5, and 5·5 months of age (mean body weight ± s.e. = 63 ± 2·5, 89 ± 3·7, 104 ± 3·7, 142 ± 3·6, 169 ± 8·1 kg, respectively), or were sham castrated at 5·5 months of age (171 ± 2·9 kg body weight) to serve as intact controls specific to this age group. Blood samples were collected at 15- to 30-min intervals from 2 h before until 8 h after treatment, with further samples collected at 10 and 12 h on day 0, and on days 1, 2 and 3, and weekly from days 7 to 35 after treatment. Following castration, peak plasma cortisol responses were significantly greater in 5·5-month-old castrates than intact calves, and all calves castrated at a younger age had reduced peak responses, with the greatest reduction in 1·5-month-old castrates. Overall, the integrated cortisol responses for the first 3 h after castration were three-fold greater in 5·5-month-old castrates than intact calves. While the integrated cortisol responses were reduced by proportionately 0·46 and 0·35 in 1·5- and 4·5-month-old castrates, the lower responses observed in 2·5- and 3·5- month-old castrates were not significantly different from the 5·5-month-old castrates. The integrated cortisol responses for the next 9 h after castration were not different among treatment groups. On day 3 after castration, peak plasma haptoglobin and fibrinogen concentrations were significantly greater in 5·5-month-old castrates than intact calves, but the concentrations were markedly reduced in calves castrated at 1·5 and 2·5 months than when castrated at 5·5 months of age. On day 1, phytohaemagglutinin-induced in vitro interferon-γ production was suppressed in 5·5-month-old castrates compared with intact calves. Scrotal circumferences increased in all castrates on day 1 and 7 and were greater in 5·5- month-old castrates than intact calves, but the swelling was reduced in the 1·5- compared with 5·5-month-old castrates. The temperature differences between the core body and scrotal skin were greater on day 2 and 3 in the 1·5-month-old than all other castrates. There was no effect of castration on the overall 42-day growth rates of calves. In conclusion, the physiological stress and inflammation caused by Burdizzo castration, indicated by increased plasma cortisol, acute-phase proteins, scrotal swelling, and depressed temperature differences between the core body and scrotal skin were reduced by castrating calves at 1·5 months rather than at 5·5 months of age.