Jutten et al.1 in this issue of Neurology® examine the implications of the intrinsic variability in the measurement of cognitive and behavioral functions for designing and interpreting Alzheimer disease (AD) clinical trials. They first identified participant data from the Alzheimer’s Disease Neuroimaging Initiative that matched inclusion criteria from 2 trials of aducanumab.2 Using longitudinal cognitive data, they estimated cognitive change from baseline over 18 months using common clinical trial endpoints, the Mini-Mental State Examination, the Clinical Dementia Rating-Sum of Boxes, and the Alzheimer’s Disease Assessment Scale Cognitive (ADAS-cog). With these estimates, they performed thousands of simulations to determine the distribution of treatment-placebo group differences, or average treatment effects that could be expected given that no treatment was administered. In other words, within the context of their inclusion criteria, they sought to define a range of change over time that represented noisy measurement. With the exception of the Phase 3 Study of Aducanumab (BIIB037) in Early Alzheimer’s Disease (EMERGE) trial, the average treatment effects of all of the other recent AD trials1 fell within the 95% boundaries of random variation.