Incorporation of l-[3- 14C]serine into phosphatidylserine (PS) and phosphatidylethanolamine (PE) has been studied in isolated rat hepatocytes. Ethanolamine inhibited the incorporation, indicating competition with serine in the base-exchange reaction. Choline, monomethylethanolamine, dimethylethanolamine and dimethyl-3-aminopropan-1-ol had no such effect. The observed rate of PS biosynthesis corresponded to 7–17 nmol/min per liver at 0.55 mM l-serine. The results indicate that only a small fraction ( 1 25 to 1 70 ) of the PS pool equilibrates with the base-exchange enzyme, and that decarboxylation to PE occurs preferentially from this pool. The rate of PS synthesis and decarboxylation can therefore not be calculated by methods which assume random, homogeneous labelling of the total PS pool. The apparent rate of PS decarboxylation increased approx. 4-fold when l-serine increased from 0.5 to 2.25 mM, suggesting that decarboxylation of PS to PE might be regulated by the concentration of l-serine or by the amount of PS present in the hepatocyte cell membranes. Lauric, palmitic, stearic, oleic and linoleic acid decreased the rate of PS synthesis. At 0.5 mM, lauric and palmitic acid were most inhibitory. At 1.0 mM, linoleic acid was the least inhibitory fatty acid. The saturated hexaenoic and saturated tetraenoic species of PS contained 51 and 29%, respectively, of the incorporated l-[3- 14C]serine. The combined monoene dienoic/diene dienoic fraction had the highest rate of synthesis judged by its relative specific activity. At 0.9 mM concentration, linoleic acid doubled the relative specific activity of the combined monoene dienoic/diene dienoic fraction of PS. Incorporation of l-[3- 14C]serine into molecular species of PE resembled that into PS, both in the absence and presence of linoleic acid, suggesting that the phosphatidylserine decarboxylase (EC 4.1.1.65) has a low specificity towards the fatty acid composition of PS. The results indicate that biosynthesis of PS from l-serine occurs mainly by the base-exchange with only negligible contribution from direct incorporation of phosphatidic acid or diacylglycerol. Furthermore, the deacylation-reacylation pathway seem to contribute only little to the determination of the fatty acid composition of hepatocyte PS. Active PS turnover seems to be confined to a small fraction of the PS pool.