Patients with primary refractory Hodgkin lymphoma (ref-HL) can still be salvaged with high dose chemotherapy (HDC) and autologous stem cell transplantation (auto-SCT). Outcome of patients with ref-HL is poorer than those with relapsed HL, but most studies have included patients with both relapsed and refractory diseases, and separate analyses or studies on patients with ref-HL are limited. This study aimed to evaluate the outcomes of HDC auto-SCT and impact of various prognostic factors on patients with ref-HL both at the time of primary treatment failure and also on subsequent survival at the time of failure post HDC auto-SCT. This was a retrospective, single-institution, cohort analysis using HDC and auto-SCT database, approved by the Institutional Research Advisory Counsel and Ethics Committee for identifying patients. We used Fine and Gray competing risk analysis method, regression model for outcome analysis and Kaplan-Meier method (KM) for survival. Two hundred consecutive ref-HL patients underwent HDC auto-SCT between 1996 to 2019. Median age was 22.75 years, median follow-up 106 months. Post auto-SCT, disease status was complete remission (CR), partial remission, and progressive disease in 122 (61%), 22 (11%), and 47 (23.5%) patients, respectively. KM median progression-free survival (PFS) after auto-SCT was 43.9 months (5:10 years, 49.3%:45.5%). Median overall survival (OS) was 168.6 months (5:10 years, 61.2%:56.2%). Eighty-five patients (44.5%) died - 69 (34.5%) due to disease. For both PFS and OS, multivariate analysis identified similar adverse factors. For PFS, stage III-IV at relapse (HR=1.65, P=0.045), mediastinal involvement (HR=2.01, P=0.009), and no CR after salvage chemotherapy (HR=2.2, P=0.001) as adverse factors. PFS with 0-1 (not reached), 2 (40.8), 3 adverse factors (5.4 months) were significant (p<0.001). For OS, stage III-IV at relapse (HR=1.68, P=0.045), mediastinal involvement (HR=2.52, P=0.007), and no CR after salvage chemotherapy (HR=2.15, P=0.004) were significant. OS with 0-1 (not reached), 2 (148.5) 3 adverse factors (34.4 months) were significant (p<0.001). Median OS after auto-SCT failure was 23.6 months; patients received post auto-SCT brentuximab/second SCT (not reached), other treatments (22.5 months), and supportive care (8.4) (p<0.001). OS with five risk factors, present at HDC auto-SCT failure, (stage III-IV, failure <12 months, tumor >5 cm, B-symptoms, low albumin) showed that 0-1:2:3-5 risk factors had 152:30.9:9.45 months OS (p<0.001). Ref-HL patients have encouraging survival after HDC auto-SCT and can even be salvaged after auto-SCT failure. Based on prognostic factors, survival prediction is possible. Patients who fail to respond to HDC auto-SCT may benefit from newer treatments strategies and may qualify for enrollment in clinical trials.