To ascertain long-term outcome of treatment for primary epithelial malignancies of the lacrimal gland and compare outcomes after cranio-orbital resection or after macroscopic tumor resection with radiotherapy. Comparative case series of 79 patients (49 male; 62%) treated for primary epithelial malignancies of the lacrimal gland at Moorfields Eye Hospital between 1972 and 2014. Patients were identified from clinical and pathological databases and, where available, the clinical, pathological, and imaging records reviewed. The primary outcome measures were overall survival after diagnosis, disease-free survival, and final visual acuity for patients having cranio-orbital resection (exenteration plus local bone removal), compared with macroscopic tumor resection plus radiotherapy. The mean age at presentation was 48 years (median: 50 years; range: 13-84 years), with 53 (67%) having adenoid cystic carcinoma, 15 (19%), primary adenocarcinoma, and 11 (14%) carcinoma ex-pleomorphic adenoma (malignant mixed tumor). The overall survival probability of the cohort (79 patients) was 0.59 at 5 years and 0.52 at 10 years, with 36/79 (46%) patients suffering tumor-related deaths; 14 patients died from other causes, and 4 patients were lost to follow up after the minimum follow-up period. The probability of disease-free survival at 5 years for patients with adenoid cystic carcinoma, adenocarcinoma, and malignant mixed tumor was 0.52, 0.4, and 0.64, respectively, with the comparable figures at 10 years being 0.44, 0.40, and 0.64. Most importantly, the 9 patients undergoing cranio-orbital resection and the 44 having solely macroscopic tumor resection plus radiotherapy had similar overall survival (p = 0.59) and disease-free survival (p = 0.89). Subgroup analysis of the 2 treatment modalities for patients with adenoid cystic carcinoma (8 cranio-orbital resection and 32 debulking and radiotherapy) demonstrated similar results for disease-free survival (p = 0.87). Likewise, there were no significant differences between rates of recurrences between the 2 different treatments. For the 50 patients who had eye-preserving surgery and long-term visual acuity data, the final acuity was better or equal to 0.6 logMAR (6/24 Snellen) in 25 (50%). There is no difference in either survival or tumor recurrence for lacrimal gland carcinoma treated with cranio-orbital resection, or eye-preserving tumor excision and radiotherapy. The authors, therefore, continue to advocate local resection and radiotherapy for almost all patients with primary epithelial malignancies of the lacrimal gland-this treatment having lower morbidity, causing less disfigurement, and, importantly, preserving useful vision in most patients.