Ninety preselected children, aged between 8 and 14 years, living in two rural West African (Gambian) villages, were randomly divided into three groups, matched for age and sex. One group received a placebo (lactose) tablet, one received riboflavin (5 mg) on 5 d every week, which was sufficient to correct an endemic riboflavin deficiency, and one received a multivitamin supplement (Protovit; Hoffmann La Roche), on 5 d every week, together with FeSO4 (200 mg) once weekly, and the supplements were given for 1 year. Neuromuscular tests, including arm tremor and manipulative skills, were performed on three occasions: once just before the introduction of the supplements; again 6 weeks after commencing the supplements; and again 1 year later. Venous blood samples were collected at the same time as the first two sets of neuromuscular tests. These samples were used for haematology and nutrient status indices: plasma ferritin, ascorbic acid, cyanocobalamin and pyridoxal phosphate, and erythrocyte tests for folate status, for riboflavin status (erythrocyte glutathionc reductase activation coefficient) and thiamine status (erythrocyte transketolase activation coefficient). The riboflavin in both supplements achieved a clear-cut response in biochemical status, which was dose-dependent. The pyridoxine, ascorbic acid and Fe components of the multivitamin also affected the associated biochemical indices. Although overall the arm tremor and related neuromuscular function tests did not respond significantly to the supplements, significant improvement was seen in the boys for the arm-tremor test in both the supplemented groups.