Here we employ a novel neuroanatomical normative modelling approach to index individual patterns of brain structure in patients with Alzheimer’s disease or mild cognitive impairment. By comparing patients to a large independent dataset (n>33,000), our analysis provides evidence that i) patterns of cortical thickness differ markedly between individual patients with Alzheimer’s disease, ii) these individual cortical ‘outlier’ patterns are associated with cognitive performance and CSF measures, and iii) individualised cortical atrophy measures can predict conversion from the MCI stage to diagnosed Alzheimer’s disease.