In this chapter, we introduce normative modeling as a tool for mapping variation across large neuroimaging datasets. We provide practical guidance to illustrate how normative models can be used to map diverse patterns of individual differences found within the large datasets used to train the models. In other words, while normative modeling is a method often applied to big datasets containing thousands of subjects, it provides single subject inference and prediction. We use an open-source Python package, Predictive Clinical Neuroscience Toolkit (PCNtoolkit) and showcase several helpful tools (including an interface that does not require coding) to run a normative modeling analysis, evaluate the model fit, and visualize the results.