Pericardial effusion is classified according to its onset—acute, subacute, or chronic (>3 months)—distribution (circumferential or loculated), and haemodynamic impact. Concerning the size, we propose a simple semiquantitative echocardiographic assessment: mild (<10 mm), moderate (10–20 mm), and large (>20 mm), evaluated as the largest telediastolic echo-free space in two-dimensional mode. Symptoms vary according to the speed of accumulation; slow accumulation may induce no or minor symptoms. In the presence of chronic, large pericardial effusions, appropriate tests for neoplasms, tuberculosis, and hypothyroidism should be considered. Chest computed tomography scanning is helpful in reaching an aetiological diagnosis (neoplasms, lymphomas, pneumonia, tuberculosis). High values of proteins, albumin, and lactate dehydrogenase are usually considered indicative of an exudate, as in pleural fluid, but this may not be true for pericardial fluid, and cytology has a sensitivity of only 50% for neoplasm. Mycobacterium cultures and a genome search for tuberculosis with the polymerase chain reaction in pericardial fluid are mandatory if pericardiocentesis is performed. If inflammatory signs are present, the clinical management should be that of pericarditis and a trial of non-steroidal anti-inflammatory drugs, colchicine, or low-dose corticosteroids, or a combination of these, may be tried. In about 60% of cases, the effusion is associated with a known disease, and the therapy should be targeted. When pericardiocentesis is performed in large effusions, prolonged pericardial drainage of up to 30 mL/24 h has been suggested to prevent recurrences, although evidence to support this is scarce. Prognosis is related to the aetiology, and idiopathic effusions may have a good prognosis especially if the effusion is mild to moderate.