Platinum-based doublet chemotherapy with or without bevacizumab is the standard treatment for untreated advanced nonsquamous non-small-cell lung cancer (NS-NSCLC). However, adding bevacizumab to chemotherapies other than paclitaxel-carboplatin is, though widely applied clinically, largely unjustified due to the lack of head-to-head data. We performed a Bayesian network meta-analysis (NMA) to address this important issue. Data of 8,548 patients from 18 randomized controlled trials (RCTs) receiving six treatments, including taxane-platinum (Taxane-Pt), gemcitabine-platinum (Gem-Pt), pemetrexed-platinum (Pem-Pt), taxane-platinum + bevacizumab (Taxane-Pt + B), gemcitabine-platinum + bevacizumab (Gem-Pt + B) and pemetrexed-platinum + bevacizumab (Pem-Pt + B), were incorporated into the analyses. Direct and indirect evidence of overall survival (OS) and progression-free survival (PFS) were synthesized at the hazard ratio (HR) scale and evidence of objective response rate (ORR) and serious adverse events (SAE) were synthesized at the odds ratio (OR) scale. Taxane-Pt + B showed significant advantages in OS (HR = 0.79, p < 0.001), PFS (HR = 0.54, p < 0.001) and ORR (OR = 2.7, p < 0.001) over Taxane-Pt with comparable tolerability (OR = 3.1, p = 0.08). Gem-Pt + B showed no OS benefit compared to any other treatment. No significant differences were detected between Pem-Pt + B and Pem-Pt in four outcomes. In terms of the benefit-risk ratio, Pem-Pt and Taxane-Pt + B were ranked the first and second, respectively. In conclusion, in the first-line treatment for advanced NS-NSCLC, Taxane-Pt and Gem-Pt are the most and least preferable regimens to be used with bevacizumab, respectively. Adding bevacizumab to Pem-Pt remains unjustified because it fails to improve efficacy or tolerability. In terms of the benefit-risk ratio, Pem-Pt and Taxane-Pt + B are the best and second-best treatment for this population.