BACKGROUND: The units of different preparations of botulinum neurotoxin type A (BoNT-A) have different potencies, and dosing recommendations for each product are not interchangeable. Historically, there has been debate concerning the dose-equivalence ratio that should be used in clinical practice. METHODS: Published evidence was considered to establish an appropriate dose-conversion ratio for the two main commercially available preparations of BoNT-A--Dysport (Dp) and Botox (Bx). RESULTS: Four key areas of evidence were identified: nonclinical and preclinical studies; studies exploring the diffusion characteristics and effects of complexing proteins; comparative experimental data from human studies; and clinical studies. Nonclinical data indicate that the principal reasons for differences in unit potency between the two products are dilution artefacts in the mouse assay. Use of saline as a diluent, at high dilutions, results in significant loss of potency in the Bx assay, whereas use of gelatin phosphate buffer in the Dp assay procedure protects the toxin during dilution. The published data on mouse assays show a Dp : Bx unit ratio range of 2.3-2.5 : 1 in saline and 1.8-3.2 : 1 in gelatin phosphate buffer. Data indicate that complexing proteins or size of the complex, which is highly pH sensitive, play no role in toxin diffusion and that Dp and Bx have similar diffusion characteristics when used at comparable doses. Randomized, controlled clinical studies indicate that 3 : 1 is more appropriate than 4 : 1, but the two products are not equivalent at this ratio. Comparative human experimental studies using the extensor digitorum brevis test, facial lines and anhidrotic action halo tests support dose-conversion ratios less than 3 : 1. LIMITATIONS: Data comparing dose equivalence ratios from the non-clinical setting should be extrapolated into the clinical setting with some caution. CONCLUSIONS: Dose-conversion ratios between Dp and Bx of 4 : 1 and greater are not supported by the recent literature.