In humans and rats we found that nicotine transiently modifies the blink reflex. For blinks elicited by stimulation of the supraorbital branch of the trigeminal nerve, nicotine decreased the magnitude of the orbicularis oculi electromyogram (OOemg) and increased the latency of only the long-latency (R2) component. For blinks elicited by electrical stimulation of the cornea, nicotine decreased the magnitude and increased the latency of the single component of OOemg response. Since nicotine modified only one component of the supraorbitally elicited blink reflex, nicotine must act primarily on the central nervous system rather than at the muscle. The effects of nicotine could be caused by direct action on lower brainstem interneurons or indirectly by modulating descending systems impinging on blink interneurons. Since precollicular decerebration eliminated nicotine's effects on the blink reflex, nicotine must act through descending systems. Three lines of evidence suggest that nicotine affects the blink reflex through the basal ganglia by causing dopamine release in the striatum. First, stimulation of the substantia nigra mimicked the effects of nicotine on the blink reflex. Second, haloperidol, a dopamine (D2) receptor antagonist, blocked the effect of nicotine on the blink reflex. Third, apomorphine, a D2 receptor agonist, mimicked the effects of nicotine on the blink reflex.