Biochemical examination of pleural fluid is usually done to try to identify the cause of a pleural effusion. The various analytes that have been suggested for this are reviewed and evaluated. Distinguishing whether the effusion is an exudate or transudate is a pragmatic first step. with further investigations dictated by the clinical features and these results. Total protein and lactate dehydrogenase were used first; Light's criteria were published in 1972 and since then additional markers including cholesterol, bilirubin and albumin gradient plus combinations of these have been proposed. Although combination testing does improve the sensitivity for diagnosis of an exudate. this is at the expense of specificity. Measurement of fluid to serum ratios appears to confer no advantage, and if a single test is required total protein performs as well as any. Additional tests may be useful in specific circumstances: pleural fluid pH may aid decisions over drainage of a parapneumonic effusion; glucose may indicate an effusion associated with rheumatoid arthritis; and adenosine deaminase may help with the diagnosis of tuberculous effusions.