INTRODUCTION: This study evaluated noninferiority of darbepoetin alfa versus placebo for overall survival (OS) and progression-free survival (PFS) in anemic patients with NSCLC treated to a 12.0-g/dL hemoglobin (Hb) ceiling. METHODS: Adults with stage IV NSCLC expected to receive two or more cycles of myelosuppressive chemotherapy and Hb less than or equal to 11.0 g/dL were randomized 2:1 to blinded 500 μg darbepoetin alfa or placebo every 3 weeks. The primary endpoint was OS; a stratified Cox proportional hazards model was used to evaluate noninferiority (upper confidence limit for hazard ratio [HR] < 1.15). Secondary endpoints were PFS and incidence of transfusions or Hb less than or equal to 8.0 g/dL from week 5 to end of the efficacy treatment period. RESULTS: The primary analysis set included 2516 patients: 1680 were randomized to darbepoetin alfa; 836 to placebo. The study was stopped early per independent Data Monitoring Committee recommendation after the primary endpoint was met with no new safety concerns. Darbepoetin alfa was noninferior to placebo for OS (stratified HR = 0.92; 95% confidence interval [CI]: 0.83‒1.01) and PFS (stratified HR = 0.95; 95% CI: 0.87‒1.04). Darbepoetin alfa was superior to placebo for transfusion or Hb less than or equal to 8.0 g/dL from week 5 to end of the efficacy treatment period (stratified odds ratio = 0.70; 95% CI: 0.57‒0.86; p < 0.001). Objective tumor response was similar between the groups (darbepoetin alfa, 36.4%; placebo, 32.6%). Incidence of serious adverse events was 31.1% in both groups. No unexpected adverse events were observed. CONCLUSIONS: Darbepoetin alfa dosed to a 12.0-g/dL Hb ceiling was noninferior to placebo for OS and PFS and significantly reduced odds of transfusion or Hb less than or equal to 8.0 g/dL in anemic patients with NSCLC receiving myelosuppressive chemotherapy.