Aimed at comparing diagnostic accuracy of ¹⁸F-FDG PET with PET/CT for pulmonary neoplasm, a study based on multi-center clinical trial of the diagnoses, in randomized and semi-blind ways, was executed from January 2006 to June 2007. It included 55 patients, i.e. 16 with histopathologically proved lung tumors, 16 with tuberculosis and 23 with benign lesions (inflammation, pseudotumor, granuloma, fibrosis and others). The histopathologic and clinic results were served as reference standard. Statistical significances in pulmonary nodule diagnosis between ¹⁸F-FDG PET and PET/CT were determined with 95% confidence interval obtained by ROC analysis. The ¹⁸F-FDG PET detected lung neoplasm with a sensitivity of 87.5% (14/16), a specificity of 59.0% (23/39), an accuracy of 67.3% (37/55) and a positive-likelihood ratio of 2.13. The ¹⁸F-FDG PET/CT detected lung neoplasm with a sensitivity of 93.8% (15/16), a specificity of 61.5% (24/39), an accuracy of 70.9% (39/55) and a positive-likelihood ratio of 2.43. The area under curves (AUC) of ¹⁸F-FDG PET and PET/CT were 0.803±0.068 and 0.799±0.063, respectively. It can be concluded that the diagnostic accuracy for malignant pulmonary nodules between ¹⁸F-FDG PET and PET/CT was not statistically different.