The aim of this study was to investigate the effects of different general anaesthesia techniques on immune responses in patients undergoing surgery for tongue cancer. Sixty American Society of Anesthesiologists physical status 1 or 2 patients undergoing elective reconstructive surgery for tongue cancer were randomised to three groups. Group 1 received propofol induction and maintenance (TIVA), group 2 received propofol induction and sevoflurane maintenance (MIXED) and group 3 received sevoflurane induction and maintenance (SEVO). All patients received an infusion of remifentanil. Blood samples were obtained at eight time-points: 30 minutes before induction (T0); one hour (T1), three hours (T2) and five hours (T3) after induction; at the end of the operation (T4); and 24 hours (T5), 48 hours (T6) and 72 hours (T7) after operation. The T lymphocyte subsets (including CD3(+) cells, CD3(+)CD4(+) cells and CD3(+)CD8(+)cells) and CD4(+)/CD8(+) ratio, natural killer cells and B lymphocytes were analysed by flow cytometry. All immunological indicators except CD3(+)CD8(+) cells were significantly decreased in all groups at T1~T5 compared to T0 (P <0.05). The percentages of CD3(+) cells, CD3(+)CD4(+) cells and natural killer cells, and the CD4(+)/CD8(+) ratios were significantly lower in the MIXED groups and SEVO groups but not the TIVA group at T6 as compared with T0 (P <0.05). There were minor but statistically significant differences in the percentages of CD3(+) cells, CD3(+)CD4(+) cells and natural killer cells, and the CD4(+)/CD8(+) ratios between the SEVO group and the TIVA group at T2approxT6 (P <0.05). These findings suggest that propofol has slightly less effect on cellular immune responses than sevoflurane.