Although muscle-relaxant doses of botulinum A toxin (BoNT/A) are generally lower than doses stimulating the immune system, specific antibodies are raised in a substantial number of patients. As a rule, this necessitates the termination of treatment. Therefore, a reliable determination of specific anti-BoNT/A antibodies is helpful and we introduced, for this purpose, a novel in vitro toxin-neutralizing assay based on a nerve-muscle preparation. We measured the antibody titers in four groups of subjects: Group 1 comprised 75 randomly selected patients of a total of 295 who responded to treatment with Dysport in our local clinic. Five patients, in group 2, were nonresponders. Group 3 consisted of 32 untreated volunteers and group 4 of 8 subjects immunized with a toxoid more than 10 years ago. Two of the responders had marginal titers of neutralizing antibodies, while they were present in all nonresponders. The sera of all responders were also tested for nonneutralizing antibodies by ELISA. Their occurrence, however, was of no consequence to the therapeutic success. The blood samples of volunteers were free from specific antibodies, whereas antibodies persisted in the immunized subjects for longer than a decade. Patients from various clinics who had been treated unsuccessfully with the toxin-14 patients had received BOTOX, 7 had been treated with Dysport, and 7 with both products-all had neutralizing antibodies. Whether there was an antibody response depended on the amount of toxin administered. We believe, however, the effective toxin dose can be reduced by so much as to make antibody production highly improbable.