Tel: + 1 714 246 54 13; fax: + 1 714 246 4374Ongoing investigations evaluated clinically relevant properties of BOTOX" (botulinum toxin type A) relative to other botulinum neurotoxin preparations based on the same (type A) or different (types B, C,, E and F) serotypes. The mouse Digit Abduction Scoring (DAS) assay was used to compare muscle weakening efficacy, the antigenic potential of two BOTOX" preparations was measured in rabbits, and the presence of antibodies to serotypes A and B was analysed in humans. BOTOX" and new BOTOX" produced similar degrees of dose-related muscle weakness in mice. Both preparations of BOTOX" were approximately four times more potent than Dysport". Preparations of serotypes B, C, and F also demonstrated lower potency than BOTOX", with serotype F also having a shorter duration of action. Neutralising antibodies were found in rabbits 3 months post-treatment with BOTOX", but were undetected 8 months post-treatment with new BOTOX". A high incidence of antibodies to type B was observed in individuals with no known exposure to type B toxin: highest in groups with the highest incidence of type A antibodies. The safety margin for BOTOX", calculated using DAS median effective dose (ED,) and the minimum dose producing a 10% reduction in body weight, was more than twice that of Dysport". In conclusion, each botulinum toxin serotype tested exhibited a different muscle weakening efficacy; BOTOX" consistently exhibited the greatest eficacy. Importantly, BOTOX" and Dysport exhibited markedly different efficacy and safety profiles, and should not be considered interchangeable. Antibody distribution in humans suggests that there may be immunological cross-reactivity between serotypes A and B. Eur J Neurol 6 (suppl4):S3-S10 0 Blackwell Science Ltd