Cobalamin-deficient L1210 mouse leukemia cells were propagated on [3H]folate under conditions designed to completely label the intracellular folate compounds. The latter, present almost entirely as polyglutamate forms, were converted to the corresponding monoglutamates by treatment with gamma-glutamylcarboxypeptidase; the monoglutamates were subsequently separated and identified by chromatography on Sephadex G-25. 5-Methyltetrahydrofolate accounted for approximately 70% of the total folate pool (17.8 nmol/10(9) cells), while the remainder was comprised largely of 5-formyltetrahydrofolate (13.0%), and tetrahydrofolate (9.5%). The identity of 5-methyltetrahydrofolate were further confirmed by chromatography on DEAE-cellulose and by its quantitative conversion to tetrahydrofolate after treatment with a highly purified preparation of methionine synthetase (5-methyltetrahydrofolate:homocysteine methyltransferase, EC 2.1.1.13). In contrast, cobalamin-replete cells contained only 5.3% 5-methyltetrahydrofolate, along with 30% each of the other three reduced folate compounds. These results provide direct experimental proof of the "methyl trap" hypothesis, which predicted that 5-methyltetrahydrofolate should accumulate in cobalamin-deficient cells.