Rat brain contains substantial concentrations of free malonate (192 nmol/g wet weight) but origin and biological importance of the dicarboxylic acid are poorly understood. A dietary source has been excluded. A recently described malonyl-CoA decarboxylase deficiency is associated with malonic aciduria and clinical manifestations, including mental retardation. In an effort to study the metabolic origin of free malonate, several labeled acetyl-CoA precursors were administered by intracerebral injection. [2-14C]pyruvate or [1,5-14C]citrate produced radioactive glutamate but failed to label malonate. In contrast, [1-14C]acetate, [2-14C]acetate, and [1-14C]butyrate were converted to labeled glutamate and malonate after the same route of administration. The intracerebral injection of [1-14C]-beta-alanine as a precursor of malonic semialdehyde and possibly free malonate did not give rise to radioactivity in the dicarboxylate. The labeling pattern of malonic acid is compatible with the reaction sequence: acetyl-CoA----malonyl-CoA----malonate. The final step is thought to occur by transfer of the CoA-group from malonyl-CoA to succinate and/or acetoacetate. Labeling of malonate from acetate is most effective at the age of 7 days when the net concentration of the dicarboxylic acid in rat brain is still very low. At this age, butyrate was a better precursor of malonate than acetate. It is proposed that fatty acid oxidation provides the acetyl-CoA which functions as the precursor of free brain malonate. Compartmentation of malonate biosynthesis is likely because the acetyl-CoA precursors citrate and pyruvate are ineffective.