To conduct a network meta-analysis of randomised controlled trials to determine the optimal clinical choice of first-line therapy for patients with ALK receptor tyrosine kinase (ALK) gene rearrangement non-small cell lung cancer (NSCLC).Clinical trials in patients with histologically confirmed ALK gene rearrangement NSCLC, that included ALK inhibitors as first-line therapy, were identified using database searches. A Bayesian network meta-analysis was conducted to calculate the efficacy and safety of the included first-line treatments.Nine trials with 2,407 patients were included for analyses. Lorlatinib was better than brigatinib for progression-free survival (PFS) (hazard ratio 0.79, 95% confidence interval 0.63, 0.98). In subgroup analyses, lorlatinib exhibited the highest probability of best PFS ranking in patients with or without baseline brain metastases (38% and 80%, respectively); brigatinib had the highest probability of best PFS ranking among Asian patients (47%). Alectinib offered the highest survival advantage (57% probability), while lorlatinib was likely to be the best treatment for an objective response (41% probability). Alectinib displayed the highest probability of being ranked lowest for grade ≥3 adverse events (86%).Lorlatinib was associated with the best PFS overall, and was suitable for patients with or without brain metastases. Brigatinib was associated with the best PFS in Asian patients.