The metabolism of S-methylmethionine has been studied in cultures of plants of Lemna paucicostata and of cells of carrot (Daucus carota) and soybean (Glycine max). In each system, radiolabeled S-methylmethionine was rapidly formed from labeled l-methionine, consistent with the action of S-adenosyl-l-methionine:methionine S-methyltransferase, an enzyme which was demonstrated during these studies in Lemna homogenates. In Lemna plants and carrot cells radiolabel disappeared rapidly from S-methylmethionine during chase incubations in nonradioactive media. The results of pulse-chase experiments with Lemna strongly suggest that administered radiolabeled S-methylmethionine is metabolized initially to soluble methionine, then to the variety of compounds formed from soluble methionine. An enzyme catalyzing the transfer of a methyl group from S-methylmethionine to homocysteine to form methionine was demonstrated in homogenates of Lemna. The net result of these reactions, together with the hydrolysis of S-adenosylhomocysteine to homocysteine and adenosine, is to convert S-adenosylmethionine to methionine and adenosine. A physiological advantage is postulated for this sequence in that it provides the plant with a means of sustaining the pool of soluble methionine even when overshoot occurs in the conversion of soluble methionine to S-adenosylmethionine. The facts that the pool of soluble methionine is normally very small relative to the flux into S-adenosylmethionine and that the demand for the latter compound may change very markedly under different growth conditions make it plausible that such overshoot may occur unless the rate of synthesis of S-adenosylmethionine is regulated with exquisite precision. The metabolic cost of this apparent safeguard is the consumption of ATP. This S-methylmethionine cycle may well function in plants other than Lemna, but further substantiating evidence is neeeded.