Allergic rhinitis is a common condition which, at its most severe, can significantly impair quality of life despite optimal treatment with antihistamines and topical nasal corticosteroids. Allergen injection immunotherapy significantly reduces symptoms and medication requirements in allergic rhinitis but its use is limited by the possibility of severe systemic reactions. There has therefore been considerable interest in alternative routes for delivery of allergen immunotherapy, particularly the sublingual route. The objective was to evaluate the efficacy of sublingual immunotherapy (SLIT), compared with placebo, for reductions in symptoms and medication requirements. The Cochrane Controlled Clinical Trials Register, MEDLINE (1966-2002), EMBASE (1974-2002) and Scisearch were searched, up to September 2002, using the terms (Rhin* OR hay fever) AND (immunotherap* OR desensiti*ation) AND (sublingual). All studies identified by the searches were assessed by the reviewers to identify Randomized Controlled Trials involving participants with symptoms of allergic rhinitis and proven allergen sensitivity, treated with SLIT or corresponding placebo. Data from identified studies was abstracted onto a standard extraction sheet and subsequently entered into RevMan 4.1. Analysis was performed by the method of standardized mean differences (SMD) using a random effects model. P-values < 0.05 were considered statistically significant. Subgroup analyses were performed according to the type of allergen administered, the age of participants and the duration of treatment. Twenty-two trials involving 979 patients, were included. There were six trials of SLIT for house dust mite allergy, five for grass pollen, five for parietaria, two for olive and one each for, ragweed, cat, tree and cupressus. Five studies enrolled exclusively children. Seventeen studies administered the allergen by sublingual drops subsequently swallowed, three by drops subsequently spat out and two by sublingual tablets. Eight studies involved treatment for less than 6 months, 10 studies for 6-12 months and four studies for greater than 12 months. All included studies were double-blind placebo-controlled trials of parallell group design. Concealment of treatment allocation was considered adequate in all studies and the use of identical placebo preparations was almost universal. There was significant heterogeneity, most likely due to widely differing scoring systems between studies, for most comparisons. Overall there was a significant reduction in both symptoms (SMD -0.42, 95% confidence interval -0.69 to -0.15; P = 0.002) and medication requirements [SMD -0.43 (-0.63, -0.23); P = 0.00003] following immunotherapy. Subgroup analyses failed to identify a disproportionate benefit of treatment according to the allergen administered. There was no significant reduction in symptoms and medication scores in those studies involving only children but total numbers of participants was too small to make this a reliable conclusion. Increasing duration of treatment does not clearly increase efficacy. The total dose of allergen administered may be important but insufficient data was available to analyse this factor.