Isolated epidermis of Commelina communis L. and Tulipa gesneriana L. assimilated (14)CO2 into malic acid and its metabolites but not into sugars or their phosphates; epidermis could not reduce CO2 by photosynthesis and therefore must be heterotrophic (Raschke and Dittrich, 1977). If, however, isolated epidermis of Commelina communis was placed on prelabelled mesophyll (obtained by an exposure to (14)CO2 for 10 min), radioactive sugars appeared in the epidermis, most likely by transfer from the mesophyll. Of the radioactivity in the epidermis, 60% was in sucrose, glucose, fructose, 3-phosphoglyceric acid and sugar phosphates. During a 10-min exposure to (14)CO2, epidermis in situ incorporated 16 times more radioactivity than isolated epidermal strips. Isolated epidermis of Commelina communis and Tulipa gesneriana took up (14)C-labelled glucose-1-phosphate (without dephosphorylation), glucose, sucrose and maltose. These substances were transformed into other sugars and, simultaneously, into malic acid. Carbons-1 through-3 of malic acid in guard cells can thus be derived from sugars. Radioactivity appeared also in the hydrolysate of the ethanol-insoluble residue and in compounds of the tricarboxylic-acid cycle, including their transamination products. The hydrolysate contained glucose as the only radioactive compound. Radioactivity in the hydrolysate was therefore considered an indication of starch. Starch formation in the epidermis began within 5 min of exposure to glucose-1-phosphate. Autoradiograms of epidermal sections were blackened above the guard cells. Formation of starch from radioactive sugars therefore occurred predominantly in these cells. Epidermis of tulip consistently incorporated more (14)C into malic and aspartic acids than that of Commelina communis (e.g. after a 4-h exposure to [(14)C]glucose in the dark, epidermis, with open stomata, of tulip contained 31% of its radioactivity in malate and aspartate, that of Commelina communis only 2%). The results of our experiments allow a merger of the old observations on the involvement of starch metabolism in stomatal movement with the more recent recognition of ion transfer and acid metabolism as causes of stomatal opening and closing.