Herein we describe a companion ongoing randomized phase III study in patients with advanced nonsquamous non-small-cell lung cancer (NSCLC). Patients with chemotherapy-naive advanced disease will be randomized to receive either pemetrexed 500 mg/m2 plus carboplatin area under the curve (AUC) 6 for 4 cycles followed by maintenance pemetrexed (arm A) or paclitaxel 200 mg/m2 plus carboplatin AUC 6 plus bevacizumab 15 mg/kg for 4 cycles followed by maintenance bevacizumab (arm B). Cycles are 3 weeks. The primary endpoint is progression-free survival (PFS)without grade 4 toxicity (G4PFS) and will test the hypothesis that G4PFS is superior for the pemetrexed-containing combination. This type of endpoint has been used previously in clinical trials in which survival outcomes have been shown to be similar between treatment regimens; thus, a regimen that reduces the risk of toxicity is clinically relevant, particularly in the palliative setting. The study will enroll approximately 360 patients (180 per arm), allowing for a 10% drop-out. Assuming a hazard ratio (HR) of 0.75, this study will have an 80% statistical power to detect superiority of arm A over arm B with the use of a 1-sided log-rank test and a type I error of 0.05. If the true median G4PFS for arm B is 3 months, then the HR of 0.75 equals approximately 1 month of improvement in median G4PFS for arm A. A gatekeeper strategy will be used to sequentially test PFS. This strategy will preserve the overall type I error rate when conducting statistical tests on both G4PFS and PFS.