Background: The optimal management of patients with metastatic germ cell tumor (GCT) who experience progression after first-line (FL) therapy is controversial. A risk-stratified approach was introduced in 2005. Favorable risk patients received 4 cycles of paclitaxel, ifosfamide, cisplatin (TIP); unfavorable risk patients received 2 cycles of TIP followed by 2 cycles of high-dose chemotherapy [(HDCT); carboplatin and etoposide with autologous stem cell support]. Favorable risk was generally defined as gonadal primary and relapse > 6 months after CR or marker negative PR to FL therapy; all others were unfavorable risk. The aim of this study is to present survival outcome of this risk-stratified approach at a single institution. Methods: A chart-based retrospective review was conducted on all patients with metastatic GCT who experienced progression after FL therapy from 2005 to 2013. Data were collected on baseline demographics and disease characteristics. Progression-free survival (PFS) and overall survival (OS) were calculated for favorable and unfavorable risk patients. Results: 37 patients were identified and included for analysis. Patients characteristics: Caucasian 92%; median age 27 years; gonadal primary 84%; non-seminoma, 81%. 16 (43%) favorable risk patients received TIP x 4; 21 (57%) unfavorable risk patients received TIP x 2 followed by HDCT x 2. The median time for follow-up was 16 months. Favorable risk patients: 2- and 4-year PFS of 60% and 45%; 2- and 4-year OS of 74%. Unfavorable risk patients: 2- and 4-year PFS of 41% and 27%; 2- and 4-year OS of 55% and 34%. Both IGCCCG risk classification at initial diagnosis and IPFSG risk classification at relapse had prognostic validity for OS (log rank p ≤ 0.05) but not PFS (log rank p = 0.065–0.080). Conclusions: Risk-stratified treatment of patients with metastatic GCT with progression after FL therapy may result in the best risk/benefit ratio. However, the optimal management of these patients is still not known and prospective randomized comparisons are needed.