Decision trees are a family of algorithms that use a treelike structure to mimic humans’ decision-making process. This chapter presents knowledge that is needed to understand and practice decision trees. We will first focus on the basics of decision trees. In particular, we will see how a decision tree is generated in training and used for predictions. A clear understanding will be gained for the splitting of data, selection of attributes at nodes, and the underlying information theory criteria. Next, discussions on such criteria will help us reach three classic decision tree algorithms: ID3, C4.5, and CART. These algorithms will be explained with adequate detail for implementation. After that, the common issues of decision trees, especially overfitting, will be explained. In the end, countermeasures, including common pre-pruning and post-pruning methods, will be outlined.