PURPOSE: Widely used (18)F 2'-deoxy-2'-fluoro-d-glucose (FDG) positron emission tomography (PET) can be problematic with false positives in cancer imaging. This study aims to investigate the diagnostic accuracy of a candidate PET tracer, (18)F 2',3'-dideoxy-3'-fluoro-2-thiothymidine (FLT), in diagnosing pulmonary lesions compared with FDG. MATERIALS AND METHODS: After comprehensive search and study selection, a meta-analysis was performed on data from 548 patients pooled from 17 studies for evaluating FLT accuracy, in which data from 351 patients pooled from ten double-tracer studies was used for direct comparison with FDG. Weighted sensitivity and specificity were used as main indicators of test performance. Individual data was extracted and patient subgroup analyses were performed. RESULTS: Overall, direct comparisons showed lower sensitivity (0.80 vs. 0.89) yet higher specificity (0.82 vs. 0.66) for FLT compared with FDG (both p<0.01). Patient subgroup analysis showed FLT was less sensitive than FDG in detecting lung cancers staged as T1 or T2, and those ≤2.0 cm in diameter (0.81 vs. 0.93, and 0.53 vs. 0.78, respectively, both p<0.05), but was comparable for cancers staged as T3 or T4, and those >2.0 cm in diameter (0.95 vs. 1.00, 0.96 vs. 0.88, both p>0.05). For benignities, FLT performed better compared with FDG in ruling out inflammation-based lesions (0.57 vs. 0.32, p<0.05), and demonstrated greater specificity regardless of lesion sizes. CONCLUSIONS: Although FLT cannot replace FDG in detecting small and early lung cancers, it may help to prevent patients with larger or inflammatory lesions from cancer misdiagnosis or even over-treatment.