To determine the effect of botulinum toxin type A on calcitonin gene-related peptide secretion from cultured trigeminal ganglia neurons. The ability of botulinum toxins to cause muscle paralysis by blocking acetylcholine release at the neuromuscular junction is well known. Previous studies and clinical observations have failed to demonstrate sensory changes related to botulinum toxins or the disease of botulism. Recent studies, however, have suggested that botulinum toxin type A injected into pericranial muscles may have a prophylactic benefit in migraine. This observation has renewed the debate of a mechanism of sensory inhibition mediated by botulinum toxin type A. Primary cultures of rat trigeminal ganglia were utilized to determine whether botulinum toxin type A could directly decrease the release of calcitonin gene-related peptide, a neuropeptide involved in the underlying pathophysiology of migraine. Untreated cultures or cultures stimulated with a depolarizing stimulus (potassium chloride) or capsaicin, an agent known to activate sensory C fibers, were treated for 3, 6, or 24 hours with clinically effective doses of botulinum toxin type A or a control vehicle. The amount of calcitonin gene-related peptide secreted into the culture media following the various treatments was determined using a specific radioimmunoassay. A high percentage (greater than 90%) of the trigeminal ganglia neurons present in 1- to 3-day-old cultures was shown to express calcitonin gene-related peptide. Treatment with depolarizing stimuli (potassium chloride), a mixture of inflammatory agents, or capsaicin caused a marked increase (4- to 5-fold) in calcitonin gene-related peptide released from the trigeminal neurons. Interestingly, overnight treatment of trigeminal ganglia cultures with therapeutic concentrations of botulinum toxin type A (1.6 or 3.1 units) did not affect the amount of calcitonin gene-related peptide released from these neurons. The stimulated release of calcitonin gene-related peptide following chemical depolarization with potassium chloride or activation with capsaicin, however, was greatly repressed by the botulinum toxin, but not by the control vehicle. A similar inhibitory effect of overnight treatment with botulinum toxin type A was observed with 1.6 and 3.1 units. These concentrations of botulinum toxin type A are well within or below the range of tissue concentration easily achieved with a local injection. Incubation of the cultures with toxin for 24, 6, or even 3 hours was very effective at repressing stimulated calcitonin gene-related peptide secretion when compared to control values. These data provide the first evidence that botulinum toxin type A can directly decrease the amount of calcitonin gene-related peptide released from trigeminal neurons. The results suggest that the effectiveness of botulinum toxin type A in the treatment of migraine may be due, in part, to its ability to repress calcitonin gene-related peptide release from activated sensory neurons.