This study was performed to investigate pneumothorax characteristics and association with clinical outcomes in patients with osteosarcoma treated with apatinib. We retrospectively reviewed the medical records of osteosarcoma patients treated with apatinib between January 2016 and April 2020 at three institutions. We evaluated the prevalence, healing time, recurrence, severity, clinical management, and prognosis of pneumothorax in these patients. A total of 54 osteosarcoma patients who received apatinib treatment were enrolled in this study. Among them, 14 patients had pneumothorax. There were significant differences between the patients with and without pneumothorax with regard to the cavitating rate of lung metastases (92.86 vs. 32.50%, respectively,
P
< 0.001), objective response rate (42.86 vs. 10.00%,
P
= 0.013), disease control rate (85.71 vs. 42.50%,
P
= 0.006), 4-month progression-free survival (PFS) rate (57.10 vs. 20.00%,
P
< 0.001), and median PFS (5.65 vs. 2.90 months,
P
= 0.011). Compared with pneumothorax patients treated with chest tube drainage only [non-staphylococcal enterotoxin C (SEC) group], those treated with chest tube drainage and SEC thoracic perfusion in parallel (SEC group) had a shorter pneumothorax healing time (12.00 ± 4.50 days vs. 24.00 ± 14.63 days for SEC group and non-SEC group, respectively,
P
= 0.103), a lower recurrence rate of pneumothorax (25.00% vs. 66.67%,
P
= 0.277), and a longer median PFS (5.9 months vs. 4.75 months,
P
= 0.964). however, these numerical differences for the SEC/non-SEC data did not reach statistical significance. Pneumothorax and cavitation in lung metastases may be effective prognostic markers for patients with osteosarcoma treated with apatinib. SEC may be effective for treatment of such pneumothorax patients, warranting further study.