Previous studies have shown that the anesthetic potency of organic compounds increases as a given halogen is replaced with successively larger halogens. These studies often are limited in the accuracy of determination of potency, rarely correlate potency with physical properties, and usually fail to include ether compounds. Because establishing relationships between structure and activity may shed light on anesthetic action, we studied the new anesthetic, I-537 (CHF2-O-CHBr-CF3), relative to two other ether anesthetics, I-653 (CHF2-O-CHF-CF3) and isoflurane (CHF2-O-CHCl-CF3) for both of which MAC and oil/gas partition coefficients are accurately known. The oil/gas partition coefficient of I-537 at 37 degrees C was found to be 245 +/- 6 (mean +/- SD) and the MAC in Sprague-Dawley rats 0.52 +/- 0.07%. Increasing atomic weight of the 1-ethyl halogen (i.e., F in I-653, Cl in isoflurane, and Br in I-537) progressively decreases MAC (increases potency) and increases lipid solubility. Although potency and solubility change by more than 10-fold, the product of MAC and the oil/gas partition coefficient remains essentially constant (120 +/- 11). However, this product is significantly less than that for other inhaled anesthetics, a finding which either challenges the unitary theory of narcosis or suggests that the lipid solvent classically used to model the site of anesthetic action (olive oil) is inappropriate.