Background: Standard CT is toxic and yields unsatisfactory results in advanced NSCLC. The aim of this trial was to evaluate feasibility and efficacy of BE first-line combination, followed by CT at PD. Methods: Eligibility criteria included non-squamous NSCLC stage IIIB-IV, WHO PS 0-1, no brain metastases. B 15 mg/kg i.v. day 1 of each 21-day cycle and E 150 mg p.o. daily were given until PD or unacceptable toxicity, followed by CT with gemcitabine 1250 mg/m² i.v. on days 1 and 8 and either cisplatin 80 mg/m² or carboplatin AUC 5 i.v. on day 1 (every 3 weeks) for a maximum of 6 cycles. The primary endpoint was disease stabilization rate (DSR) after 12 weeks of BE. Secondary endpoints were overall survival (OS), response rate, time to progression (TTP), DSR at 6 and 18 weeks, safety, quality of life, response to subsequent CT, molecular pathology (mutations), allele-specific polymerase chain reaction (PCR) for mutations (tumor, serum), serum proteomics and other translational analyses. Results: 103 patients (pts) were enrolled, 101 were evaluable. Under BE, DSR at week 12 was 54.5% (95% CI: 44.2-64.4%). 73 pts had at least SD (72.3%), including 1 CR (1.0%) and 17 PR (16.8%). No unexpected toxicities were observed. Median TTP under BE was 4.1 months (mo). 62 (61.4%) pts started CT; 35 received at least 4 cycles. Six PR and 32 SD were observed. At a median follow-up of 36 mo, median OS was 14 mo (95% CI: 10.7-19.0 mo). Twelve pts had an activated EGFR mutation in exons 19 or 21; their exclusion did not change overall TTP or OS, and DSR at week 12 remained 50.6%. Thirteen pts had KRAS mutation. Allele-specific PCR detected EGFR mutations in serum of > 50% of pts with tumor mutations. PCR also detected identical EGFR mutations in 66% of tumors tested by sequencing method. Conclusions: The BE regimen followed by CT at PD is feasible in advanced non-squamous NSCLC with low toxicity and promising activity. OS is similar to the observed OS using up-front standard CT regimens.