Objectives: Our aim was to review the clinical records from children with large pericardial effusions of inflammatory origin presenting to a tertiary referral centre over the last 21 years, with emphasis on their clinical presentation, management and outcome. Background: The common identifiable causes of pericardial effusion in children include prior cardiac surgery, bacterial pericarditis, malignancy, and connective tissue disorders. In a significant number of children, however, despite extensive investigation, it is not possible to identify a clear aetiology. A viral cause is often considered, though rarely confirmed. The clinical course of such large idiopathic pericardial effusions in children has not been extensively reported. Methods and results: We reviewed retrospectively the records of all patients seen between 1981 and 2001 with large pericardial effusions of inflammatory origin requiring drainage, excluding the effusions related to cardiac surgery or malignancy. We found 31 patients fulfilling our criterions for study. They could be divided into three groups, with 15 patients having no specific identifiable aetiology despite extensive investigation, 12 patients having evidence of bacterial pericarditis, and four with a probable immunologic disorder. Fever was present in only eight patients (53%) in the idiopathic group. All patients in the other groups had fever. Except for fever and the resultant tachycardia, it was not possible to distinguish on clinical grounds, nor on the presence or otherwise of cardiac tamponade, between those with idiopathic aetiology and those with bacterial infection. Of the patients with presumed bacterial pericarditis, five (42%) had both positive blood and pericardial fluid cultures, three (25%) had positive blood cultures, while a further three patients (25%) had only positive pericardial fluid cultures. All patients required drainage of the pericardial effusion, either under echocardiographic guidance or surgically. None of the patients died. The hospital stay was significantly shorter for those with idiopathic as opposed to bacterial pericarditis. Of those with an idiopathic aetiology, six required readmission due to recurrence of the pericardial effusion, with four patients requiring further surgical drainage. No patients required readmission with a bacterial or immunologic aetiology. No patient developed constrictive pericarditis after a median follow-up of 22 months. Conclusion: Patients with large idiopathic pericardial effusion had relatively few constitutional symptoms as compared with their gross echocardiographic findings. Those with bacterial pericarditis had more urgent need for treatment. Patients with pericardial effusion of inflammatory origin, when treated appropriately, had an excellent outcome with no mortality or development of constrictive pericarditis.