Background Many general thoracic surgeons are learning robotic pulmonary resection. Methods We retrospectively compared results of completely portal robot lobectomy with 4 arms (CPRL-4) against propensity-matched controls and results after technical changes to CPRL-4. Results In 14 months, 168 patients underwent robotic pulmonary resection: 7 had metastatic pleural disease, 13 had conversion to open procedures, and 148 had completion robotically (106 lobectomies, 26 wedge resections, 16 segmentectomies). All patients underwent R0 resection and removal of all visible lymph nodes (median of 5 N2, 3 N1 nodal stations, 17 lymph nodes). The 106 patients who underwent CPRL-4 were compared with 318 propensity-matched patients who underwent lobectomy by rib- and nerve-sparing thoracotomy. The robotic group had reduced morbidity (27% vs 38%; P  = .05), lower mortality (0% vs 3.1%; P  = .11), improved mental quality of life (53 vs 40; P  < .001), and shorter hospital stay (2.0 vs 4.0 days; P  = .02). Results of CPRL-4 after technical modifications led to reductions in median operative time (3.7 vs 1.9 hours; P  < .001) and conversion (12/62 vs 1/106; P  < .001). Technical improvements were addition of fourth robotic arm for retraction, vessel loop to guide the stapler, tumor removal above the diaphragm, and carbon dioxide insufflation. Conclusions The newly refined CPRL-4 is safe and yields an R0 resection with complete lymph node removal. It has lower morbidity, mortality, shorter hospital stay, and better quality of life than rib- and nerve-sparing thoracotomy. Technical advances are possible to shorten and improve the operation.