This comprehensive guide provides an introduction to online learning and online convex optimization: the theory of making decisions when data are revealed over time. It develops the core ideas of regret minimization through a unified treatment of classic and modern algorithms, and is the first textbook to cover parameter-free algorithms. Written in an informal but mathematically precise style, it emphasizes intuition, rigorous guarantees, and the connections between algorithms that are often presented separately. Readers will learn how online learning tools support adaptive prediction, sequential decision-making, confidence guarantees, game-theoretic reasoning, and learning in changing environments. Concluding exercises and 'History Bits' tracking the development of methods and ideas in the field of online learning enhance each chapter. With no prior knowledge of convex analysis required, the book is suited to graduate students, researchers, and practitioners working in machine learning, optimization, statistics, theoretical computer science, operations research, and related fields.