The object of the present paper was to evaluate the diagnostic accuracy of ABR using ipsilateral masking in patients suffering from multiple sclerosis. Fourteen patients were examined. Each patient with normal audiogram underwent two ABR recording sessions during a follow-up period of thirty months. The ABRs were recorded under two experimental conditions applying standard testing procedures and ipsilateral masking, respectively. In 5 patients magnetic resonance imaging was also available. The findings were compared with a control group, composed of 20 normal subjects. Standard ABR testing gave evidence of a pathological value in 50% of the cases while the ABRs proved to be altered in 64% under ipsilateral masking. The outcomes were less favourable than those obtained with somatosensory or visually evoked potentials. The conclusions drawn confirm the importance of ABR in the follow-up of multiple sclerosis. On the other hand, ipsilateral masking does not appear to provide any significant advantages due to poor amelioration of sensitivity and doubtful specificity of the observed alterations.