Retro-orbital granuloma is a rare and devastating component of granulomatosis with polyangiitis (GPA). Current medical treatment protocols are falling short, and outcomes are poor. The aim of the study was to investigate the frequency, clinical features, and treatment outcomes of retro-orbital granuloma in patients with GPA. This is a retrospective, multi-centre study, which involves GPA cohorts from five different clinics. Data were extracted from patient charts including history, physical examination, radiological-laboratory-histological findings, and treatment protocols. Major clinical outcome measures were changes in the volume of the granuloma on comparative MRI, and visual acuity on repeated ophthalmologic examinations. Among 141 GPA patients, nine (five females and four males) were diagnosed with a retro-orbital granuloma. Median duration of disease was 8 years. Proptosis and diplopia were the dominant presenting symptoms (77 %), followed by orbital pain (55 %). Three out of nine patients had isolated retro-orbital granulomas, without other organ involvement of GPA. Five patients received conventional pulse steroid and pulse cyclophosphamide (CYC) as the first-line remission induction therapy. Four of these patients had progressive disease, and a regression in granuloma size was observed in one patient using this regimen. Two patients were already receiving immunosuppressants when they were diagnosed with retro-orbital granuloma. Six patients had been treated with RTX as the second-line remission induction therapy. None of these patients had progression following RTX therapy. Three patients underwent orbital decompression surgery. The indication for the decision for surgery was either progressive loss of vision or intractable pain. Standard first-line chemotherapy (CYC and steroids) was ineffective against retro-orbital granuloma associated with GPA. RTX could be an alternative in these cases. Surgical intervention may help to decrease the morbidity. Further prospective studies with greater patient numbers are needed to test the clinical efficiency of RTX as a first-line remission induction chemotherapy.