An "anti-oxidant cocktail" consisting of betacarotene, vitamins B6, C, E, zinc, and selenium or corresponding placebos were given for one y as daily dietary supplements to 45 elderly residents of a nursing home. Initially, the serum TBA reactant levels were higher (2.7 +/- 0.7 mumol/L) than those of an ad hoc control group of healthy younger adults (2.3 +/- 0.6 mumol/L), p less than 0.01. After three mo supplementation, the levels among the verum elderly had decreased to 2.2 +/- 0.6 mumol/L, and they remained at this lower level until the end of the study period, whereas the placebo group showed no change. A significant inverse correlation (r = -0.428, p less than 0.01) existed between the concentrations of serum TBA reactants and whole blood selenium (B-Se), but only B-Se levels above 200 micrograms/L were associated with a decrease in serum lipid peroxides. Serum alpha-tocopherol concentration also correlated inversely with serum TBA reactants but this correlation (r = -0.273, p less than 0.76) was not as strong as that of B-Se. Deficient vitamin B6 status, in biochemical terms, was observed in 25% of the elderly; a daily supplement of 2 mg B6 fully cured all cases of deficiency. The verum group improved slightly in several psychological tests, whereas subjects on placebo remained unchanged or deteriorated during the follow-up period. Clinical amelioration among the verum subjects was reported by the nurses; no toxic side effects were reported. In conclusion, the elderly benefited biochemically and clinically of dietary antioxidant supplements.