OBJECTIVES: This research was conducted to understand key reasons why the G-BA came to a different benefit rating than IQWiG during oncology HTAs. METHODS: Searching the G-BA and IQWiG homepages, oncology HTAs between 1^st of January 2011 and 31^stof December 2013 have been identified. Assessments for which the G-BA attested a benefit rating divergent from the one proposed by IQWiG have been analyzed to reveal the key reasons for difference. RESULTS: In the observed time frame, 18 HTAs of oncology products have been conducted by IQWiG and G-BA in Germany. IQWiG and G-BA were aligned in their benefit ratings in 14 reports (78%). During four assessments (22%) the G-BA came to a different conclusion than IQWiG. These are the assessments of crizotinib, eribulin, pertuzumab and the resubmission for vandentanib. One reason for different ratings observed in three assessments is that the G-BA looked at time-adjusted analyses to correct for different lengths of treatment in the trial arms. The extend of benefit that is extracted from the same data also potentially differs. This is exemplified by the case of vandentanib, where the G-BA was more convinced by the data on “time to worsening of pain” than IQWiG and attested a “minor benefit” (as compared to “no added benefit”). In the case of eribulin, G-BA and IQWiG agreed on benefit based on survival data. Due to the potential side effects however, G-BA demoted to “minor”, while IQWiG initially demoted to “no added benefit”. CONCLUSIONS: The translation of clinical evidence into a benefit rating and weighing of positive versus negative effects is highly complex. We see that G-BA and IQWiG can come to different conclusions. In all cases it is important that the manufacturer shows evidence against the specified comparator. The benefit rating plays an important role in the reimbursement amount negotiations.