OBJECTIVE The nucleoside analogue 3'-deoxy-3'-(18)F-fluorothymidine (FLT) has recently been introduced for imaging cell proliferation with positron emission tomography (PET). We prospectively evaluated whether FLT uptake reflects proliferative activity as indicated by the Ki-67 index in non-small cell lung cancer (NSCLC), in comparison with 2-deoxy-2-(18)F-fluoro-D-glucose (FDG). METHODS A total of 18 patients with newly diagnosed NSCLC were examined with both FLT PET and FDG PET. PET imaging was performed at 60 min after each radiotracer injection. Tumour lesions were identified as areas of focally increased uptake, exceeding background uptake in the lungs. For semi-quantitative analysis, the maximum standardised uptake value (SUV) was calculated. Proliferative activity as indicated by the Ki-67 index was estimated in tissue specimens. Immunohistochemical findings were correlated with SUVs. RESULTS The sensitivity of FLT and FDG PET for the detection of lung cancer was 72% and 89%, respectively. Four of the five false-negative FLT PET findings occurred in bronchiolo-alveolar carcinoma. The mean FLT SUV was significantly lower than the mean FDG SUV. A significant correlation was observed between FLT SUV and Ki-67 index (r = 0.77; p < 0.0002) and for FDG SUV (r = 0.81; p < 0.0001). CONCLUSIONS The results of this preliminary study suggest that, compared with FDG, FLT may be less sensitive for primary staging in patients with NSCLC. Although FLT uptake correlated significantly with proliferative activity in NSCLC, the correlation was not better than that for FDG uptake.