Data about the feasibility and safety of thoracoscopic surgery under non-intubated anesthesia and regional block are limited. In this prospective study, 57 consecutive patients scheduled for thoracoscopic surgery were enrolled. Patients were sedated with dexmedetomidine and anesthetized with propofol and remifentanil. Ropivacaine was used for intercostal nerve and paravertebral block. Lidocaine was used for vagal block. The primary outcomes were mean arterial pressure (MAP), heart rate (HR), oxygen saturation, and end-tidal carbon dioxide partial pressure (ETCO2) at T0 (pre-anesthesia), T1 (immediately after laryngeal mask/nasopharyngeal airway placement), T2 (immediately after skin incision), T3 (10 min after opening the chest), T4 (end of surgery), and T5 (immediately after laryngeal mask/nasopharyngeal airway removal). One patient required conversion to intubation, 15 developed intraoperative hypotension, and two had hypoxemia. MAP at T0 and T5 was higher than at T1–T4; MAP at T3 was lower (P<0.05 vs other time points). HR at T0 and T5 was higher (P<0.05 vs other time points). ETCO2 at T2 and T3 was higher (P<0.05 vs other time points). Arterial pH, PCO2, and lactic acid at T1 differed from values at T0 and T2 (P<0.05). The Quality of Recovery-15 (QoR-15) score at 24 h was lower (P<0.05). One patient experienced dysphoria during recovery. Thoracoscopic surgery with regional block under direct thoracoscopic vision is a feasible and safe alternative to conventional surgery under general anesthesia, intubation, and one-lung ventilation.