BACKGROUND: Flares of eczema are attributed to many factors, often with minimal scientific evidence. OBJECTIVES: Systematically to search, summarize and critically appraise the scientific evidence to support the roles of individual 'flare factors' in eczema. METHODS: We searched Medline from 1966 until 20 April 2005 to identify relevant articles for inclusion in this review. No language restrictions were imposed. All study designs were included and were ranked according to the strength of evidence. Experimental and provocation studies were restricted to those using a double-blind design. We included randomized controlled trials if they were provocation studies. Meta-analysis was not possible due to differences in study populations and methodology. The studies are therefore described qualitatively. RESULTS: The roles of foodstuffs (13 studies), house dust mite (three), other aeroallergens (two), seasonality (two), bacterial infections (one), textiles (three), detergents (one), sunlight (one) and stress (two) were assessed in different study populations, using a variety of study designs. All studies were performed on selected groups and only four were longitudinal in design. Collectively, these studies provide some evidence that certain foods, house dust mite, stress and seasonal factors are relevant causes of disease worsening in certain subgroups with eczema. No good evidence could be found to support the role of detergents, textiles and irritants in causing worsening of eczema. CONCLUSIONS: Despite anecdotal lists in textbooks and review articles, very little good evidence exists for 'flare factors' in eczema. The focus of all of the included studies was on disease worsening rather than clinically relevant flares. Studies of longitudinal design are required to clarify the roles of these and other putative flare factors in eczema.