Current FDA guidelines for AD clinical trials include demonstrating efficacy on both a cognitive and a global measure to ensure the clinical meaningfulness of drug effects. The ADAS-Cog is the standard cognitive assessment tool in clinical trials. However it has well-known ceiling effects that limit its ability to measure cognitive changes early in the disease. This presents an important challenge to research efforts aimed at intervening early in the disease process with disease-modifying therapies. We present an empirical characterization of the ADAS-Cog and subscales; identify the important subscales for MCI and early AD, and evaluate additional cognitive and functional measures. Novel composite outcome measures based on cognitive and functional measures are also discussed. *Data used in the preparation of this talk were obtained from the ADNI database (adni.loni.ucla.edu).