The possibility that the phospholipid-N-methyltransferases from yeast are capable of acting upon a phospholipid substrate, localized in a different membrane than in which the enzymes reside ('trans-catalysis' hypothesis), was investigated using cho2 and opi3 gene disruptant strains, which are defective in phosphatidylethanolamine transferase (PEMT) and phospholipid methyltransferase (PLMT), respectively. When cell homogenates or microsomes of the two disruptant strains are mixed, the combined methyltransferase activity, measured as the incorporation of [(3)H]methyl label from S-adenosyl methionine, exceeds that expected based on the separate activities of PEMT and PLMT. The increased incorporation implies that monomethylphosphatidylethanolamine generated by PEMT becomes available for PLMT, as evidenced by increased synthesis of dimethylphosphatidylethanolamine and phosphatidylcholine. The kinetics of the cooperativity suggest a collision-based process, enabling either transport of substrate or 'trans-catalysis'.