We studied two brothers (J.R. and M.R.) with the cobalamin D variant of congenital methylmalonic aciduria-homocystinuria, whose previously reported lack of megaloblastic anemia conflicted with current concepts of cobalamin's role in DNA synthesis and the "methyltetrahydrofolate (MTHF) trap" hypothesis. Both subjects were indeed hematologically normal, although J.R. had a mean corpuscular volume of 96 fl. However, both demonstrated abnormalities in the deoxyuridine suppression test. J.R. had an abnormal suppression value of 21.0% (normal less than 10%) that was correctable by adding hydroxocobalamin or folic acid in vitro but not MTHF. M.R. had normal suppression (8.9%), but demonstrated worsening (18.6%) when MTHF was added. J.R.'s classical deoxyuridine suppression pattern of cobalamin deficiency thus supports the trap hypothesis. However, his lack of comparable morphological changes suggests that impaired de novo thymidylate synthesis and the trap hypothesis, though valid, may not fully account for the megaloblastic maturation accompanying cobalamin deficiency. Equally noteworthy was the deleterious effect of MTHF on M.R.'s marrow, suggesting its potential usefulness as an in vitro "stress test" for latent cobalamin abnormality.