1. A comparative study of the actions of structurally diverse allosteric modulators on mammalian (human alpha 3 beta 2 gamma 2L) or invertebrate (Drosophila melanogaster Rdl or a splice variant of Rdl) recombinant GABA receptors has been made using the Xenopus laevis oocyte expression system and the two electrode voltage-clamp technique. 2. Oocytes preinjected with the appropriate cRNAs responded to bath applied GABA with a concentration-dependent inward current. EC50 values of 102 +/- 18 microM; 152 +/- 10 microM and 9.8 +/- 1.7 microM were determined for human alpha 3, beta 1 gamma 2L, Rdl splice variant and the Rdl receptors respectively. 3. Pentobarbitone enhanced GABA-evoked currents mediated by either the mammalian or invertebrate receptors. Utilizing the appropriate GABA EC10, the EC50 for potentiation was estimated to be 45 +/- 1 microM, 312 +/- 8 microM and 837 +/- 25 microM for human alpha 3, beta 1 gamma 2L, Rdl splice variant and Rdl receptors respectively. Maximal enhancement (expressed relative to the current induced by the EC10 concentration of GABA where this latter response = 1) at the mammalian receptor (10.2 +/- 1 fold) was greater that at either the Rdl splice variant (5.5 +/- 1.3 fold) or Rdl (7.9 +/- 0.8 fold) receptors. 4. Pentobarbitone directly activated the human alpha 3 beta 1 gamma 2L receptor with an EC50 of 1.2 +/- 0.03 mM and had a maximal effect amounting to 3.3 +/- 0.4 fold of the response evoked by the EC10 concentration of GABA. Currents evoked by pentobarbitone were blocked by 10-30 microM picrotoxin and potentiated by 0.3 microM flunitrazepam. Pentobarbitone did not directly activate the invertebrate GABA receptors. 5. 5 alpha-Pregnan-3 alpha-ol-20-one potentiated GABA-evoked currents mediated by the human alpha 3 beta 1 gamma 2L receptor with an EC50 of 87 +/- 3 nM and a maximal enhancement of 6.7 +/- 0.8 fold of that produced by the GABA EC10 concentration. By contrast, relatively high concentrations (3-10 microM) of this steroid had only a modest effect on the Rdl receptor and its splice variant. 6. A small direct effect of 5 alpha-pregnan-3 alpha-ol-20-one (0.3-10 microM) was detected for the human alpha 3 beta 1 gamma 2L receptor (maximal effect only 0.08 +/- 0.01 times that of the GABA EC10). This response was antagonized by 30 microM picrotoxin and enhanced by flunitrazepam (0.3 microM). 5 alpha-Pregnan-3 alpha-ol-20-one did not directly activate the invertebrate GABA receptors. 7. Propofol enhanced GABA-evoked currents mediated by human alpha 3 beta 1 gamma 2L and Rdl splice variant receptors with EC50 values of 3.5 +/- 0.1 microM and 8 +/- 0.3 microM respectively. The maximal enhancement was similar at the two receptor types (human 11 +/- 1.8 fold; invertebrate 8.8 +/- 1.4 fold that of the GABA EC10). 8. Propofol directly activated the human alpha 3 beta 1 gamma 2L receptor with an EC50 of 129 +/- 10 microM, and at a maximally effective concentration, evoked a current amounting to 3.5 +/- 0.5 times that elicited by a concentration of GABA producing 10% of the maximal response. The response to propofol was blocked by 10-30 microM picrotoxin and enhanced by flunitrazepam (0.3 microM). Propofol did not directly activate the invertebrate Rdl splice variant receptor. 9. GABA-evoked currents mediated by the human alpha 3 beta 1 gamma 2L receptor were potentiated by etomidate (EC50 = 7.7 +/- 0.2 microM) and maximally enhanced to 8 +/- 0.8 fold of the response to an EC10 concentration of GABA. By contrast, the Rdl, or Rdl splice variant forms of the invertebrate GABA receptor were insensitive to the positive allosteric modulating actions of etomidate. Neither the mammalian nor the invertebrate receptors, were directly activated by etomidate. 10. delta-Hexachlorocyclohexane enhanced GABA-evoked currents with EC50 values of 3.4 +/- 0.1 microM and 3.0 +/- 0.1 microM for the human alpha 3 beta 1 gamma 2L receptor and the Rdl splice variant receptor respectively. The maximal enhancement was 4.5