Otitis represents the second most common infection of the upper respiratory tract, its treatment being the most common cause for prescribing antibiotics in the United States. A large number of antimicrobials, especially beta-lactams and macrolides, are generally used for treating acute otitis media (AOM) in paediatric patients, owing to their antibacterial spectrum including the main aetiological pathogens. Efficacy, safety and compliance of Cefaclor were compared with those of other antibiotics in the treatment of paediatric AOM in a meta-analysis of randomized controlled trials published between 1981 and 2004. Overall, evaluations were performed on 24 studies (Medline/PubMed, keywords "Cefaclor and otitis") which proved eligible (jadad score > or = 1); sixteen out of the 24 studies were multicentre, seven were double-blind. Mostly, the comparator agent was a beta-lactam, in four and three cases it was a macrolide or the association trimethoprim-sulfamethoxazole, respectively. Efficacy and safety were end-points of all studies whereas only 9 studies evaluated compliance. For the majority of studies (16/24) Cefaclor was administered for 10-day course. The analysis was based on a 2 x 2 contingency table with classification by treatment and number of improvements/cures, side-effects, and compliance of individual studies. The global estimate of the effective treatments was obtained with the weighted mean of the log OR (Odd Ratio) according to Mantel-Haenszel and associated confidence intervals (CI) at 95%. All the calculations were performed using SAS v.8. Chi-square test was performed. Clinical efficacy evaluation, number of improvements/cures, did not evidence a statistically significant difference among Cefaclor and comparators (86.8% vs 88.7%; Odds Ratio 0.77, IC 0.61/0.94). In the Cefaclor-treated patients, adverse events were observed in a statistically significant lower percentage compared to other antibiotics: 13.3% vs 19.4% (P < 0.0001), diarrhoea and gastro-intestinal disturbances being the most frequently observed. Compliance was observed in a similar proportion in the two patient groups, Cefaclor and comparators (Cefaclor 88.1; comparators 91.1%) and the slight difference was not statistically significant (Odds Ratio 0.77, IC 0.39-1.15). The present meta-analysis proves that in the treatment of paediatric AOM Cefaclor exhibits a clinical efficacy equal to other antibiotics usually employed in this setting, similar compliance but superior safety.