The stereoselectivity of certain anesthetics is currently thought to be inconsistent with lipid theories of narcosis. The EC(50)-values of etomidate enantiomers for tadpole narcosis are now examined as a function of octanol/water partition coefficients, and enhancement factors for predicted over experimental EC(50) values are determined from a calibration curve for non-selective narcosis. The unfavored S-(-)-enantiomers of etomidate and two analogues surprisingly still fulfill the Meyer-Overton rule. The R(+)-enantiomers of etomidate and four structural analogues are up to 34-fold more active than expected. The non-chiral anesthetic, propofol, is 8-fold more active than expected. It is concluded that there may be two pathways to tadpole narcosis: enhanced narcosis involving specific receptor binding sites and non-selective narcosis corresponding to the Meyer-Overton rule and operating on the lipid/protein interface.