The level of plasma total homocysteine (tHcy), long known to be B vitamin dependent, has recently been shown to be inversely associated with plasma selenium (Se) concentration in human subjects. We therefore, chose to investigate the interaction between Se, tHcy and B vitamins in a double-blind, placebo-controlled trial where 501 healthy UK elderly volunteers were randomly allocated to receive 100, 200, or 300 lug Se/day as high-Se-yeast, or placebo-yeast for 6 months. Plasma Se, tHcy, folate, vitamin B-12, pyridoxal-5'-phosphate (PLP) and its catabolite, 4-pyridoxic acid. were measured in all participants at baseline and in samples from the placebo, 100 and 300 mu g Se/day groups, at follow-up. At baseline, Se was inversely correlated with tHcy but only in males (p < 0.001). Before supplementation, tHcy concentration was significantly lower in the highest compared to the lowest Se tertile in males (p < 0.05), and in females when folate concentrations were also in the top tertile (p < 0.05). The effect of folate, PLP and vitamin B-12 concentrations on plasma tHcy correlated with Se concentration at baseline. After 6 months of Se supplementation, only Se concentration had changed significantly. Supplementation with Se does not affect tHcy concentration in the UK elderly population.