One hundred sixty-two white-blood-cell scans were retrospectively reviewed to determine the sensitivity and specificity of the test for pulmonary and pleural infection. All scans were performed 18-24 hr after injection of indium-111 oxine-labeled autologous or donor cells. Pulmonary activity was graded on a scale of 0-4: 0 = equal to soft tissue; 1 = greater than soft tissue but less than rib; 2 = equal or greater than rib but less than liver; 3 = equal or greater than liver but less than spleen; 4 = equal to spleen. Activity was also characterized as being focal or diffuse. The white-blood-cell scan findings were correlated with the clinical diagnosis on the basis of physical examination, laboratory results, chest radiographs, clinical course, and pathologic studies when available. As pulmonary activity increased from grade 1 to 4, sensitivity declined from 93% to 14% and specificity increased from 64% to 100%. The sensitivity and specificity of focal uptake were 31% and 89% vs 62% and 74% for diffuse pulmonary activity. Making a distinction between focal and diffuse activity did not improve the specificity of low grades of pulmonary activity. The white-blood-cell scan can be very sensitive or very specific for pulmonary or pleural infection, depending on the criteria selected for a positive scan.