We are developing an affordable smartphone tool designed to inexpensively, accurately and efficiently diagnose dementia during its pre-dementia stage. Tablets and smartphones now allow for the capture of real time and real world background information. Our team has already developed a suite of tablet and smartphone based tests, the CNTookit and iBOCA, which capture not only traditional testing metrics but also real world information. We propose to conduct clinical pilot testing of our tablet and smartphone based apps, which collects passive and active data, to create a novel database of normal, MCI, AD and non-AD dementia patients’ performance across their naturalistic environments as well as their changes over time. We will use machine-based analysis to identify the optimal tests and environmental factors and create an adaptive smartphone app that relies on the most efficient series of tests for identifying dementia.