Pulmonary infarction needs to be distinguished from several diseases. In its differentiation from myocardial infarction the electrocardiogram is a very helpful investigation, but it can be difficult to distinguish pulmonary infarction from pneumonia. Evidence of either a potential source of embolus or of a recent infection of the respiratory tract may facilitate the diagnosis, and the chest radiograph may sometimes be helpful, but signifi cant numbers of patients remain in whom a clear-cut distinction cannot readily be made on clinical or radiological grounds. Many serum enzyme determinations have been introduced for the investigation of cardiac and hepatic disease?for example, alkaline phosphatase, alanine and aspartate amino transferase, lactate dehydrogenase and its isoenzymes, isocitrate dehydrogenase, etc.?and the possible value of some of these assays for differentiating between pulmonary infarction and pneumonia has been investigated by various workers?for example, Nikkil? (1959), Wacker and Snodgrass (I960), Wacker, Rosenthal, Snodgrass, and Amador (1961), Stevens and Burdette (1964). So far no single biochemical test has been proposed as a reliable means of differentiating between the two conditions, but combinations of tests have been advocated. Nikkil? (1959) tentatively suggested that measurement of aspartate aminotransferase (G.O.T.), glucose-6-phosphate dehydrogenase, and alkaline phosphatase might prove a helpful triad of investigations, and Wacker and Snodgrass (1960) and Wacker et al. (1961) advocated serial determinations of lactate dehydrogenase (L.D.), G.O.T., and serum bilirubin. In 17 patients diagnosed clinically (12) or at necropsy (5) as having had pulmonary infarction, with or without embolism, L.D. was elevated, whereas G.O.T. remained normal, and in 12 out of the 17 patients serum bilirubin was abnormal at some stage of the illness. Wacker et al. (1961) also applied this combination of tests to seven other patients who subsequently came to necropsy (five with myocardial infarction and two with bacterial pneumonia) ; the patients with myocardial infarction showed abnormal G.O.T. and L.D. values, whereas the patients with pneumonia had normal serum enzyme levels. They therefore proposed that the finding of an elevated L.D. and bilirubin, in association with a normal G.O.T., favoured a diagnosis of pulmonary infarction in patients with a suggestive clinical history. Stevens and Burdette (1964) were unable to confirm the value of the combination of tests advocated by Wacker et al. (1961) ; the only finding consistently observed by Stevens and Burdette (1964) in patients with pulmonary infarction was an elevation of L.D., and they suggested that an elevated L.D. might be a requirement for the laboratory diagnosis of pulmonary infarc tion. However, even this single test is probably not uniformly reliable, as Wr?blewski, Ruegsegger, and LaDue (1956) and Kirkeby and Prydz (1959) have reported normal L.D. values in patients with pulmonary infarction.