Eighteen infants with infantile spasms were given sodium dipropylacetate at a dosage of 20mg/kg/day. They were clinically examined before treatment, and again after one to three years of therapy. The short-term clinical response was excellent in four patients, good in eight, poor in four and there was no change in two. At follow-up, two patients were clinically normal, but 10 had severe and six had moderate mental retardation. Seven patients still had residual seizures. Since these results do not differ significantly from those obtained with hormonal treatment, the authors suggest using sodium dipropylacetate (which has less frequent and less severe side-effects than adreno-corticotropic hormone) as the only initial drug, and to use hormonal treatment only in unresponsive patients.