After they had been ‘isolated’ by a low pH treatment which destroyed all the other cells of the epidermis, stomatal guard cells retained the capacity to produce malate when presented with fusicoccin as a stimulus to opening. For a given level of stomatal opening, the amount of malate in epidermis with isolated stomata was practically the same as in intact epidermis. It is concluded that malate transport to and from the subsidiary cells is not an essential part of the stomatal mechanism. Changes in CO2 concentration over the range 0–100 ppm affected neither stomatal aperture nor malate levels in epidermis of Commelina communis. This suggests that carboxylation of phosphenolpyruvate plays no major part in the formation of malate in guard cells.