BACKGROUND: Emergence agitation after sevoflurane anesthesia in children can be prevented by midazolam. Alternative splicing of the GABAA receptor changes with age. Therefore, we hypothesized that alternative splicing of the gamma2 subunit affects the GABA current when applying sevoflurane and midazolam. METHODS: We performed the whole-cell patch clamp technique on human embryonic kidney 293 cells that were transfected with alpha1beta2gamma2L or alpha1beta2gamma2S. The concentration-response relations were recorded for midazolam and sevoflurane, and the co-application responses were measured at concentrations of 1.5 nM, 15 nM and 300 nM of midazolam and 0.5%, 2.0% and 4.0% of sevoflurane. Each GABA current was compared with that produced by 5 microM of GABA. RESULTS: The concentration-response relationships for midazolam and sevoflurane were dose-dependent without any differences between the alpha1beta2gamma2L and alpha1beta2gamma2S subtypes. 1.5 nM and 15 nM of midazolam did not significantly enhance the current after treatment with 0.5% sevoflurane for both subtypes. The current after treatment with 2.0% sevoflurane was enhanced by 1.5 nM midazolam for the alpha1beta2gamma2S subtype, but not for the alpha1beta2gamma2L subtype. In the case of 2.0% sevoflurane with 15 nM of midazolam, and 4.0% sevoflurane with 300 nM of midazolam, the GABA currents were significantly enhanced for both subtypes. CONCLUSIONS: These results show that the difference in the gamma2 subunit cannot explain the emergence agitation after sevoflurane anesthesia in children in vitro. This suggests that co-application of sevoflurane and midazolam enhances the GABA current according to the alternative splicing of the gamma2 subunit and the concentration of both drugs.