Various sterols and related steroids were tested for their ability to influence ethanol-induced electrolyte leakage fron Hordeunm vulgare roots. Cholesterol had the greatest influence and, depending on concentration, it stimulated or inhibited the loss of electrolyte. Cholesterol, however, was ineffective if the roots were pretreated with ethanol. These data suggest that sterols protect rather than restore membrane structure. First, modifications in the cholesterol perhydrocyclopentanophenanthrene ring system suggest that at least one double bond is required for membrane activity. Second, increasing the bulkiness of the C1 side chain of cholesterol, as shown with campesterol, stigmasterol, and sitosterol, decreased its activity. Apparently for maximum effectiveness the sterol molecule should have a relatively flat configuration. Third, the C3-hydroxyl group is required for membrane activity since cholesteryl methyl ether, cholest-5-ene-3g-thiol and cholesteryl halogens were without activity. Exception to the foregoing rule was cholestane which was slightly active but which has neither a C3-hydroxyl group nor a double bond in the ring system. In previous publications it was reported that exogenously applied cholesterol, and certain other higher plant sterols, modified the permeability of plant cells (9,11,15). To explain the action of sterols on plant membranes, a hypothesis which incorporated the results of animal and phospholipid film experiments (5-7, 13, 16, 21, 26, 27) was proposed. Briefly, it was suggested that cholesterol, or other sterols, insert themselves into the phospholipid layer of the membrane, with the C.-hydroxyl group interacting by weak ion-dipole-and hydrogen-bonding with the polar nitrogenous base moiety of the phospholipid. In this model, the C, side chain of cholesterol would align itself along the nonpolar terminal of the phospholipid, filling the phospholipid cavity and in turn stabilizing the lipid configuration. According to this hypothesis, the Chydroxyl group of the sterol molecule is an important functional group. This hypothesis would explain why only the free sterols were active and not the steryl esters and steryl glycosides (9,11), all of which are common higher plant constituents. However, sterol analogs with a small polar group in the C. posi- tion, other than a hydroxyl, have never been tested. The foregoing hypothesis of sterol action also assumes that for effectiveness, the sterol molecule must have a flat configuration similar to that of cholesterol (7,26). This requirement was confirmed partially with sterols that differed in branching in the C17 side chain (9, 1 1), but in this respect sterols with modifications in the perhydrocyclopentanophenanthrene ring system have not been studied.For purposes of studying the mechanism of sterol action on plant membranes, cholesterol can be taken as the basic molecule, and this molecule can be divided into three sites of possible physiological importance (Fig. 1). The cholesterol nucleus consisting of ring A, B, C, and D with...