Aducanumab recently underwent two large phase III clinical trials that were stopped prematurely by the sponsor Biogen. One trial was trending positive while the other showed no benefits from aducanumab. Post hoc analyses led the sponsor to assert that there was a sufficient efficacy signal to justify a new drug application as a treatment for Alzheimer's disease. The sponsor claimed that subsets of participants receiving sufficiently high doses of aducanumab demonstrated benefits in both trials. In contrast, we identified alternative accounts for the apparent drug benefits in post hoc subgroups that are unrelated to dose effects. Biomarker data were consistent with target engagement, but no evidence was presented to correlate biomarker changes to cognitive benefits. Our analysis supports the conduct of a third, phase III trial with high-dose aducanumab. Aducanumab's efficacy as a treatment for the cognitive dysfunction in Alzheimer's disease cannot be proven by clinical trials with divergent outcomes.