This chapter presents the basics of reinforcement learning (RL) and, based on that, introduces value-based RL as one of the two major categories of RL algorithms. For this goal, the basic RL concepts, including Markov decision process and essential RL terms, like environment, state, action, value, reward, and policy, will be explained first. Next, we will have an RL demonstration using simple virtual environments. Then, the Bellman equation, which casts the basis for the ”learning” or improvement of an RL learning agent, will be discussed. This includes the Bellman equation’s mathematical formulations, deduction, and use. After going through the basics, we will start the introduction to value-based RL algorithms with an overview of popular RL algorithms. Following that, typical value-based RL algorithms including two time-difference algorithms, i.e., Q-learning and Sarsa, and one episode-update algorithm, i.e., Monte Carlo, will be introduced. Details will be provided to cover both theories and guidelines for the implementation of these algorithms.