Vandetanib [vascular endothelial growth factor (VEGF) receptor/epidermal growth factor receptor/RET inhibitor] has shown improvements in progression-free survival (PFS) in advanced non-small cell lung cancer in three randomized phase II studies: vandetanib versus gefitinib (study 3), docetaxel +/- vandetanib (study 6), and carboplatin-paclitaxel and/or vandetanib (study 7). In study 7, vandetanib monotherapy was inferior to carboplatin-paclitaxel. We performed an exploratory retrospective analysis of the relationship between baseline circulating VEGF concentrations and PFS. Mean baseline VEGF levels were determined by ELISA from two baseline samples of plasma (163 of 168 patients, study 3; 65 of 127, study 6) or serum (144 of 181, study 7). High baseline VEGF values were above the immunoassay reference range for healthy subjects; low baseline VEGF values were within the range. Patients with low baseline VEGF had a lower risk of disease progression with vandetanib versus gefitinib [hazard ratio (HR), 0.55; 95% confidence interval (95% CI), 0.35-0.86; P = 0.01] or vandetanib 100 mg/d + docetaxel versus docetaxel (HR, 0.25; 95% CI, 0.09-0.68; P = 0.01). High VEGF patients had a similar risk of disease progression with vandetanib monotherapy versus gefitinib (HR, 1.03; 95% CI, 0.60-1.75; P = 0.92) or vandetanib 100 mg/d + docetaxel versus docetaxel (HR, 0.95; 95% CI, 0.25-3.61; P = 0.94). In study 7, low VEGF patients had a similar risk of disease progression with vandetanib monotherapy 300 mg/d versus carboplatin-paclitaxel (HR, 0.80; 95% CI, 0.41-1.56; P = 0.51); high VEGF patients progressed more quickly (HR, 1.60; 95% CI, 0.81-3.15; P = 0.17). These analyses suggest that low baseline circulating VEGF may be predictive of PFS advantage in patients with advanced non-small cell lung cancer receiving vandetanib versus gefitinib or vandetanib + docetaxel versus docetaxel. Moreover, patients with low VEGF levels may have a similar outcome with either vandetanib monotherapy or carboplatin-paclitaxel.