Background: Radiotherapy (RT) involving craniospinal irradiation (CSI) is important in the initial treatment of medulloblastoma. At recurrence, the re-irradiation options are limited and associated with severe side-effects. Methods: For pre-irradiated patients, patients with re-irradiation (RT2) were matched by sex, histology, time to recurrence, disease status and treatment at recurrence to patients without RT2. Results: A total of 42 pre-irradiated patients with RT2 were matched to 42 pre-irradiated controls without RT2. RT2 improved the median PFS [21.0 (CI: 15.7–28.7) vs. 12.0 (CI: 8.1–21.0) months] and OS [31.5 (CI: 27.6–64.8) vs. 20.0 (CI: 14.0–36.7) months]. Concerning long-term survival after ten years, RT2 only lead to small improvements in OS [8% (CI: 1.4–45.3) vs. 0%]. RT2 improved survival most without (re)-resection [PFS: 17.5 (CI: 9.7–41.5) vs. 8.0 (CI: 6.6–12.2)/OS: 31.5 (CI: 27.6–NA) vs. 13.3 (CI: 8.1–20.1) months]. In the RT-naïve patients, CSI at recurrence improved their median PFS [25.0 (CI: 16.8–60.6) vs. 6.6 (CI: 1.5–NA) months] and OS [40.2 (CI: 18.7–NA) vs. 12.4 (CI: 4.4–NA) months]. Conclusions: RT2 could improve the median survival in a matched cohort but offered little benefit regarding long-term survival. In RT-naïve patients, CSI greatly improved their median and long-term survival.