PURPOSE: To assess positron emission tomography (PET) with 2-[fluorine-18]-2-deoxy-D-glucose (FDG) in the differentiation of recurrent bronchogenic carcinoma from fibrosis after therapy. MATERIALS AND METHODS: Any patient treated for bronchogenic carcinoma who had a residual chest radiographic abnormality was eligible. Forty-three patients (mean age, 63.5 years) participated. Chest radiographs and thoracic computed tomographic scans helped localize the abnormality prior to PET. Semiquantitative analysis was performed on FDG PET images with calculated standardized uptake ratios (SURs). Sensitivity, specificity, and confidence intervals for recurrent disease were determined. RESULTS: Thirty-five patients had recurrent or persistent tumor (median SUR, 7.6; range, 1.9-18.7). Eight patients had fibrosis but no evidence of disease (SUR, 1.6; range, 0.6-2.4). The sensitivity for detecting recurrent tumor (SUR > 2.5) was 97.1%, and specificity was 100%. The SUR for recurrent tumor was statistically significantly higher than for fibrosis (P = .0001). CONCLUSION: FDG PET accurately helps differentiate recurrent bronchogenic carcinoma from fibrosis.