Fey RIII (CD16) expression of neutrophil granulocytes was measured in 156 patients by means of fluorescence-labeled antibodies with a flow cytometer. Results were compared with (1) 400-cell manual differential count; (2) left shift flagging on hematology analyzers; (3) absolute neutrophil count; and (4) acute-phase protein levels. Asynchrony was noted between neutrophil CD16 expression and microscopically defined neutrophil stage, particularly in heavily left-shifted samples, which made it impossible to reliably enumerate immature neutrophils on the basis of CD16 expression. According to receiver operating characteristics, the absolute count of CD16-negative neutrophils was highly discriminatory for detection of left shift, with an area under the curve (AUC) of 0.842 ± 0.03 (SE) and maximum efficiency of 81% ± 3%, but absolute neutrophil count was not significantly inferior (0.821 ± 0.03 and 76% ± 3%).Some investigators question the clinical value of detection of left shift or band cells.1-5 ; others consider left shift a valuable sign for a neutrophil reaction, particularly in neonatal or pediatric patients, who may not demonstrate an increase in total neutrophil count or a humoral response of acute-phase proteins.6-8 However, both automated and manual methods of detecting left shift have serious shortcomings. Most hematology analyzers report flags designed to indicate left shift, but instrument evaluations have shown the efficiency of the flagging algorithms to be insufficient, 9,10 especially for detection of band cells.11 Lack of an adequate reference method for enumeration of band cells or immature granulocytes is one problem. Manual differential counting is characterized by high imprecision, subjectivity of the examiner, and varying definition of band cells.2,4 ' 5,12 Even after uniform band identification criteria were defined, band results remained poorly reproducible. 13 In an effort to find alternative methods toFrom the Central Laboratory, Wilhelminenspital, Vienna, Austria.M a n u s c r i p t submitted April 2, 1996; revision accepted September 4,1996.Address reprint requests to Dr Hubl: Central Laboratory, Wilhelminenspital, Montleartstrasse 37, A-1171 Vienna, Austria.STKS and SE9000 flagging demonstrated efficiency of 76% ± 3% and 81% ± 3%, respectively. For detection of acute-phase response, absolute neutrophil count (AUC, 0.836 ± 0.04; maximum efficiency, 80% ± 4%) outperformed both quantitative neutrophil CD16 expression (0.760 ± 0.05; 75% ± 4%) and absolute CD16-negative neutrophil count (0.757 ± 0.05; 71% ± 4%); absolute band count performed similarly (0.853 ± 0.04; 79% ± 4%) and showed high efficiency at high sensitivity and specificity. Efficiency of analyzer flagging for detection of acute-phase response was not superior to absolute neutrophil count (STKS, 77% ± 4%; SE9000, 78% ± 4%). In conclusion, the diagnostic value of measuring neutrophil CD16 expression was generally similar to that of less complicated analytes.