To investigate the impact of treatment lag in infantile spasms (IS) on treatment response, occurrence of later epilepsy, and long-term cognition and behavior in patients with one single etiological entity, we examined 18 patients with Down syndrome (DS) and earlier IS retrospectively (follow-up period of 32-180 months with a mean of 85.1 months), and determined their history and present condition, in terms of previously mentioned items. There was a statistically significant correlation between treatment lag and lag to cessation of spasms (R=0.55, P=0.02), developmental quotient (DQ) (R=-0.75, P=0.003), and score of autistic features (AF) (R=0.57, P=0.04). Moreover we found that the later the response to treatment of IS, the lower was the DQ (R=-0.86, P=0.001) and the higher was the score of autistic features (R=0.5, P=0.06). A long duration of spasms also determined a low DQ (R=-0.93, P<0.0001) and a high score of autistic features (R=0.66, P<0.01). All patients with persistent epilepsy (n=5) had had a treatment lag of over 2 months. Conversely, for all children treated within 2 months (n=8) spasms ceased within 3 months of treatment and none of them had later epilepsy. This group of patients with a treatment lag of less than 2 months had earlier treatment response (P=0.002), higher DQ (P=0.004) and lower score of autistic features (P=0.006). The data stress the importance of a short treatment lag in view of mental development and prevention of later epilepsy and autistic features, and raise the question of antiepileptogenic effect in this specific condition.