Patient satisfaction with botulinum toxin treatment is a key success factor in aesthetic procedures and is governed by the interaction of numerous variables. Duration of effect is important because it influences retreatment intervals as well as affecting cost and convenience to the patient. In order to review the evidence on the duration of benefit associated with various commercial formulations of botulinum toxin, MEDLINE was searched using the following terms: 'botulinum' and 'duration'/'retreatment' (limits: 'clinical trials,' 'meta-analyses,' 'English'). I also searched my existing reference files, reference lists of identified articles, and meeting/conference abstracts to ensure completeness. The focus was on clinical medicine and aesthetic trials. To be eligible for the analysis, studies had to include efficacy assessments at multiple timepoints. To estimate duration of benefit, the following outcomes were examined and summarized: responder rates, mean wrinkle severity scores at various timepoints (with or without changes from baseline), and relapse rates. Duration at both repose and maximum attempted muscle contraction was considered when provided. Where possible, duration was assessed by formulation and dose. The initial search yielded 164 articles. Of these, 35 included an adequate measure of duration in aesthetic indications. The majority of these (22) were on the glabellar area. Study designs and endpoints were highly heterogeneous, and duration of effect varied between studies. Several studies with the BOTOX Cosmetic (onabotulinumtoxinA; Allergan, Inc., Irvine, CA, USA) formulation of botulinum toxin type A (BoNTA) included relapse rates, defined conservatively as return to baseline levels of line severity for two consecutive visits approximately 30 days apart (at repose and maximum contraction). In these studies, duration of effect ranged from 3 to 5 months in female patients and from 4 to 6 months in male patients. Individual patients had longer durations of response. Across all studies providing relapse rates, most patients relapsed by 6 months. In studies assessing patient satisfaction, satisfaction remained high throughout the duration of the studies ( approximately 4 months). With the Dysport formulation (abobotulinumtoxinA, clostridium botulinum type A toxin-hemagglutinin complex; Ipsen Biopharm Ltd, Wrexham, England), retreatment intervals were estimated at a mean of 3.9 months (median = 3.3 months). These results were consistent with responder rates from another Dysport study in which the active treatment differed from placebo at 3 but not 4 months. A single comparative study demonstrated that the proportion of patients relapsing at week 16 was 23% (95% CI 11.5, 41.6) in the BOTOX Cosmetic group as compared with 40% (95% CI 25.2, 60.1) in the Dysport group. Myobloc (rimabotulinumtoxinB, botulinum toxin type B; Solstice Neurosciences, Inc., South San Francisco, CA, USA) was associated with shorter durations of action (2-3 months). Data from facial areas other than the glabella, although more limited, supported a duration of at least 3-4 months. The addition of BoNTA to dermal fillers or light/laser therapy appeared to increase the degree of effect. Repeated BoNTA treatments may prolong duration of effect or potentiate the effect. In conclusion, patients can expect treatments to last > or =3 months but often as many as 4-5 months depending on the facial area, dose, and formulation. Additional research should help clarify the impact of age, baseline rhytid severity, patient sex, repeated treatments, and combination treatment on longevity of effect.