For simple versus simple hypothesis testing, the famous Neyman-Pearson lemma (1933) provides the best fixed sample size test procedure that has minimum type II error among all the tests that control type I error at some prescribed level. However, it is well-known that even such a best test cannot guarantee control of type II error while retaining the type I error control as long as the sample size is fixed. Abraham Wald (1943) extended the Neyman-Pearson lemma for sequentially observed data by introducing the Sequential Probability Ratio Test (SPRT) that simultaneously controls type I and type II error probabilities at some prefixed levels. In this talk, I will be discussing this test procedure along with boundary conditions, average sample size, the optimality property of SPRT and truncated SPRT. If time permits, I will discuss how to extend this idea for testing k (>2) simple-vs-simple hypotheses with the goal of selecting the correct hypothesis. This problem is known as Multi-hypothesis sequential probability ratio test (MSPRT).