In the face of an exponential increase in published biomedical studies, dermatologists frequently turn to review articles in order to keep abreast of important developments in the treatment of skin diseases. Traditional review articles have recently been criticized on the basis of their incompleteness and susceptibility to bias. Such biases can be minimized by employing a systematic approach to gathering, combining and interpreting the evidence of treatment efficacy. Using eight predetermined quality criteria, we compared the quality of 10 traditional review with one systematic review of treatments for postherpetic neuralgia, which were identified from the Medline database for 1992-96. None of the 10 traditional review articles satisfied all eight criteria: one satisfied five, two satisfied four and the rest satisfied two or fewer criteria. There was a wide variation in the recommendations of the authors for the treatment of postherpetic neuralgia, often based on anecdotal evidence and clinical experience. On the other hand, the systematic review fulfilled seven of the eight quality criteria, failing only to discuss future directives. Furthermore, treatment recommendations were made solely on the basis of randomized controlled trials, which are considered to be the gold standard for measuring the benefits of any intervention. The variation in quality and treatment recommendations of traditional reviews is worrying. Systematic reviews should be encouraged in dermatology because they provide a summary of evidence of the effects of dermatological treatments, which has been derived using explicit methods widely accepted within science.