INTRODUCTION: In the randomized CeTeG/NOA‐09 trial, combined CCNU/TMZ was superior to TMZ therapy regarding overall survival (OS) in newly diagnosed patients with MGMT‐methylated glioblastoma. Using modified RANO criteria, however, progression‐free survival (PFS) and pseudoprogression rates were similar in both arms. Exploring the hypothesis of undetected pseudoprogressions being accountable for this discrepancy, we analyzed patterns of postprogression survival (PPS) and MRI features at RANO‐defined progression. METHODS: 86 CeTeG/ NOA‐09 patients with progression (modified RANO criteria) and MRI evaluable for standardized T1 and FLAIR volumetry at baseline and progression were included. Patients were further subdivided to short PPS (< 24 months) or long PPS (> 24 months) and a PPS/PFS ratio was calculated. RESULTS: In the CCNU/TMZ arm, long PPS patients (n=10) tended to a shorter PFS (median 7.3 months) than short PPS patients in the same arm (n=33, 14 months, p=0.089, logrank test) and long PPS patients in the TMZ arm (n=9, 12.7 months, p=0.21). The mean PPS/PFS ratio in the long PPS group was markedly higher in the CCNU/TMZ arm (5.8) compared to the TMZ arm (3.3, p=0.043, Mann‐Whitney test). Patients with long PPS of the CCNU/TMZ arm showed a nonsignificant tendency to a stronger volumetric increase in T1 enhancement (mean delta 6184,86 vs. 697.5 mm3) and FLAIR‐T1‐enhancement (mean delta 42671 vs 16860 mm3) at progression as compared to long PPS patients of the TMZ arm. CONCLUSION: Combining a substantially increased PPS/PFS ratio (long OS despite particularly short PFS according to RANO) with indications for increased contrast enhancement and FLAIR volume at progression, the patients with long PPS in the CCNU/TMZ arm appear to differ from those in the TMZ arm. These observations support the hypothesis that this group (∼25% of CCNU/TMZtreated patients) contained patients with pseudoprogression undetected by modified RANO criteria.