Abstract The outcome following surgery for patients with primary lung neuroendocrine tumors at metastatic stage remains poorly characterized. We conducted a retrospective national study including patients with metastatic lung neuroendocrine tumors at diagnosis. We performed a safety study to evaluate major morbidity and mortality of surgical resection of the primary tumor and compared patients in the operative to the nonoperative group. A total of 155 patients were included: 41 in the operative group and 114 in the nonoperative group, median age was 64 years. Metastases were mainly located in the liver (74.2%) and the bone (49.7%). The primary endpoint was met as the rate of major complications was 4.9% and there was no postoperative mortality. In the operative group 42.5% of patients had improvement of their pulmonary symptoms versus 14.4% in the nonoperative group. The median overall survival was not reached in the operative group and was 4.3 years (95% CI [3.5;4.9]) in the nonoperative group (univariate analysis, HR = 0.42 95% CI [0.23–0.77], p = .002). After multivariate analysis, only an ECOG‐PS ≥1 (vs. 0, HR = 2.44, 95% CI [1.46;4.07], p = .001) and >1 metastatic site (vs. 1; HR = 1.83, 95% CI [1.06;3.16], p = .030) remained significantly associated with overall survival. The resection of the primary tumor was not significantly associated with overall survival (HR = 0.63, 95% CI [0.32;1.24], p = .183). In conclusion, surgery of primary lung neuroendocrine tumors at metastatic stage is a safe option that should be considered in selected patients in order to improve symptoms with a view to improving their quality of life. Larger studies are warranted to evaluate the impact of surgery on survival.