Objective: To systematically assess the effectiveness of systemic cyclosporin in patients with severe atopic eczema. Study design: Systematic review and meta-analysis of controlled and uncontrolled trials. Electronic (MEDLINE, Cochrane databases) and hand search of published work. Independent standardized assessment of eligibility and data abstraction by two reviewers. Methods: For the qualitative review data on study design, study population, methodology, results, tolerability and methodological quality was independently extracted by two reviewers. Qualitatively homogeneous studies were pooled using a random-effects model. The mean relative change in objective disease severity was chosen as the main outcome measure for the quantitative analysis. Publication bias was explored by regressing treatment effect on sample size. Sensitivity analysis included meta-regression of study-specific covariates (inclusion of children, study type, concomitant topical therapy, study quality). Results: Fifteen studies including 602 patients met the eligibility criteria. In all studies analysed, cyclosporin consistently decreased the severity of atopic eczema. Twelve studies appeared homogeneous enough to be pooled. After 2 weeks of treatment we found a dose-related response with a pooled mean decrease in disease severity of 22% (95%-CI 8-36%) under low-dose cyclosporin (3 mg/kg) and 40% (95%-CI 29-51%) at dosages ≤4 mg/kg. After 6-8 weeks the relative effectiveness was 55% (95%t;-CI 48-62%). Conclusions: Due to evidence of publication bias the quantitative results need to be interpreted with caution. Effectiveness of cyclosporin is similar in adults and children, but tolerability might be better in children. As data to adequately evaluate the long-term effectiveness and safety of cyclosporin in patients with atopic eczema are unavailable, long-term registries are encouraged.