We report an analysis of the results obtained in our first 100 studies with indium-111-oxine labelled leukocyte scintigraphy, a diagnostic technique which has recently become available for clinical evaluation within Australia. We used this technique to assess patients with suspected sepsis or inflammation after other commonly used investigations had failed to confirm a diagnosis. Four patient subgroups were evaluated: fever of unknown origin; suspected abdominal or postoperative sepsis; suspected active inflammatory bowel disease; and suspected sepsis or inflammation of bones or joints. The course of all patients was followed for at least three months to establish the accuracy of the technique. The leukocyte labelling procedure took 90 min and imaging was carried out typically 3-6, 24 and occasionally 48 h after reinjection of the labelled autologous leukocytes. In one patient labelling of leukocytes was unsuccessful. In the remaining 99 studies the overall sensitivity of leukocyte scintigraphy was 88% (36 of 41 patients with a proved inflammatory or infective disease focus had positive scan findings); and the specificity was 95% (55 of 58 cases with no proved disease focus had normal scan findings). This series supports the use of this method as the imaging procedure of choice in nuclear medicine for the evaluation of suspected acute sepsis (symptoms less than four weeks' duration), of inflammatory bowel disease and of suspected infections involving appendicular bones which contain no active bone marrow. It is also a useful secondary scintigraphic procedure, after gallium-67-citrate scintigraphy, in patients with suspected infective disorders of more than four weeks' duration.