Event-related potentials (ERP) were assessed in 70 school-aged children with a diagnosis of asymptomatic iron deficiency anaemia (IDA), based on low haemoglobin and either low serum ferritin or high free erythrocyte protoporphyrin levels. The IDA subjects were randomized into treatment and placebo groups of 35 cases each, and compared with a normal control group of 30 age- and gender-matched healthy subjects without iron deficiency. Further haematological and ERP assessment was carried out after 3 months, during which time the active group received iron supplementation with 10 mg ferrous sulphate, together with vitamin C, malic acid and folic acid. Pre-treatment, both IDA groups had prolonged P300 latencies in comparison with the non-IDA controls (p<0.01). The proportion of cases with distorted wave appearance was more than twice as high in the IDA groups as in the non-IDA controls, although intergroup differences did not reach statistical significance. After treatment, the active treatment IDA group showed a significant increase in haemoglobin levels and shortening in P300 latencies. After treatment, neither value was statistically different from non-IDA controls. There was a decrease in the number of cases with abnormal waveforms in the active treatment group, compared with an increase in the number within the placebo group (P = 0.002). Testing of ERP shows promise as a non-invasive, sensitive and objective marker for assessing cognitive impairment in children with IDA.