Several of the two-pore domain K⁺ channels have been shown to be activated by clinically relevant concentrations of volatile (Patel et al. 1999) and gaseous (Gruss et al. 2004) general anaesthetics. Consequently, they have been proposed to play a role in inducing the general anaesthetic state. We have been studying the TASK like channels with the aim of identifying the critical regions necessary for anaesthetic modulation. It has been shown that both the TASK-1 and TASK-3 channels are potentiated by halothane. However we have found that the volatile anaesthetic chloroform (1.98 mM) causes an activation of TASK-3 (82 ± 11%; n = 17) in contrast to the inhibition that is observed on TASK-1 (27 ± 5%; n = 5) (errors given as s.e.m.). This difference can be exploited to identify the relevant regions responsible for chloroform activation. Furthermore, if we hypothesise that halothane and chloroform share an overlapping site of action then this approach may provide information about the halothane modulatory site. Previously the halothane binding site on TASK-3 was proposed to be the 243-248 VLRFLT region at the end of the fourth transmembrane (TM) domain (Talley & Bayliss, 2002). This region is highly conserved between TASK-3 and TASK-1, with only the lysine residue present in TASK-3 being replaced by a methionine in TASK-1. We investigated this residue using a point mutation to change this region from TASK-3 to TASK-1. We found, however, that the activation caused by chloroform on this L247M mutant (84 ± 10%; n = 4) was not significantly different (P < 0.01) to the wildtype channel. Due to the high sequence identity between TASK-1 and TASK-3 we were able to construct a series of chimeras between the channels to identify the region that was critical for chloroform activation. We found that the intracellular loop between TM2 and TM3 was one of the key determinants for the chloroform activation. In a parallel study (Andres-Enguix et al. in preparation) we have recently cloned the Lymnaea stagnalis K⁺ channel responsible for the I_K(AN) current (Franks & Lieb, 1988). A sequence comparison between the TASK-1, TASK-3 and the Lymnaea channels was used to identify individual residues which are important for anaesthetic activation.