1. The gamma-aminobutyric acid (GABA)-modulatory and GABA-mimetic actions of etomidate at mammalian GABA(A) receptors are favoured by beta2- or beta3- versus beta1-subunit containing receptors, a selectivity which resides with a single transmembrane amino acid (beta2 N290, beta3 N289, beta1 S290). Here, we have utilized the Xenopus laevis oocyte expression system in conjunction with the two-point voltage clamp technique to determine the influence of the equivalent amino acid (M314) on the actions of this anaesthetic at an etomidate-insensitive invertebrate GABA receptor (Rdl) of Drosophila melanogaster. 2. Complementary RNA-injected oocytes expressing the wild type Rdl GABA receptor and voltage-clamped at -60 mV responded to bath applied GABA with a concentration-dependent inward current response and a calculated EC50 for GABA of 20+/-0.4 microM. Receptors in which the transmembrane methionine residue (M314) had been exchanged for an asparagine (RdlM314N) or a serine (RdlM314S) also exhibited a concentration-dependent inward current response to GABA, but in both cases with a reduced EC50 of 4.8+/-0.2 microM. 3. Utilizing the appropriate GABA EC10, etomidate (300 microM) had little effect on the agonist-evoked current of the wild type Rdl receptor. By contrast, at RdlM314N receptors, etomidate produced a clear concentration-dependent enhancement of GABA-evoked currents with a calculated EC50 of 64+/-3 microM and an Emax of 68+/-2% (of the maximum response to GABA). 4. The actions of etomidate at RdlM314N receptors exhibited an enantioselectivity common to that found for mammalian receptors, with 100 microM R-(+)-etomidate and S-(-)-etomidate enhancing the current induced by GABA (EC10) to 52+/-6% and 12+/-1% of the GABA maximum respectively. 5. The effects of this mutation were selective for etomidate as the GABA-modulatory actions of 1 mM pentobarbitone at wild type Rdl (49+/-4% of the GABA maximum) and RdlM314N receptors (53+/-2% of the GABA maximum) were similar. Additionally, the modest potentiation of GABA produced by the anaesthetic neurosteroid 5alpha-pregnan-3alpha-ol-20-one (Rdl = 25+/-4% of the GABA maximum) was not altered by this mutation (RdlM314N = 18+/-3% of the GABA maximum). 6. Etomidate acting at beta1 (S290)-containing mammalian GABA(A) receptors is known to produce only a modest GABA-modulatory effect. Similarly, etomidate acting at RdlM314S receptors produced an enhancement of GABA but the magnitude of the effect was reduced compared to RdlM314N receptors. 7. Etomidate acting at human alpha6beta3gamma2L receptors is known to produce a large enhancement of GABA-evoked currents and at higher concentrations this anaesthetic directly activates the GABA(A) receptor complex. Mutation of the human beta3 subunit asparagine to methionine (beta3 N289M found in the equivalent position in Rdl completely inhibited both the GABA-modulatory and GABA-mimetic action of etomidate (10-300 microM) acting at alpha6beta3 N289Mgamma2L receptors. 8. It was concluded that, although invertebrate and mammalian proteins exhibit limited sequence homology, allosteric modification of their function by etomidate can be influenced in a complementary manner by a single amino acid substitution. The results are discussed in relation to whether this amino acid contributes to the anaesthetic binding site, or is essential for transduction. Furthermore, this study provides a clear example of the specificity of anaesthetic action.