Comparative studies of botulinum neurotoxin preparations to date have generally examined 2 preparations at prespecified dose ratios in relatively homogeneous groups of patients under controlled study conditions. It is unclear whether the differences in adverse-event rates that have been noted under these controlled conditions can be generalized to the broader population of cervical dystonia patients, who are treated with a wider range of doses in a variety of settings.We conducted a systematic review and analysis of the published literature to compare rates of dysphagia and dry mouth in studies of botulinum neurotoxin products.We searched the MEDLINE, EMBASE, Biosis, SciSearch, JICST (Japan Science and Technology Center), and Pascal databases from 1985 through 2006 using the terms cervical dystonia, spasmodic torticollis, and botulinum toxin for original English-language studies of Botox, Dysport, or Myobloc in the treatment of cervical dystonia (or spasmodic torticollis) that documented adverse events by treatment or patient. Studies that involved patients with various types of dystonias or movement disorders were included as long as adverse events were reported separately for those with cervical dystonia. Rates of dysphagia with the original preparation of Botox were considered separately from those with the current preparation of Botox.Seventy published articles were included in the analysis (30 Botox, 24 Dysport, 3 Botox + Dysport, 11 Myobloc, 2 Botox + Myobloc). Mean total doses per treatment ranged from 60 to 374 U for Botox, 125 to 1200 U for Dysport, and 579 to 19,853 U for Myobloc. Botox was associated with a significantly lower rate of dysphagia than Dysport, with mean dysphagia rates of 10.5% for original Botox, 8.9% for current Botox, and 26.8% for Dysport (both, P0.05). Myobloc was associated with dry mouth (3.2%-90.0%) in 9 of 13 studies, but this adverse event was not reported in a sufficient number of studies of botulinum toxin type A preparations (Botox, n = 2; Dysport, n = 6) to permit statistical comparison. In the weighted analysis, the duration of effect differed between botulinum neurotoxin products (current BotoxMyoblocoriginal BotoxDysport; all, P0.001), but only 43 (61.4%) of the 70 studies reported duration, and the definitions varied.The results of this analysis indicate differences in adverse-event rates between botulinum neurotoxin preparations, suggesting that use of these products should be based on their individual dosing, efficacy, and safety profiles.