Bayesian statistical methods are nowadays broadly applied in data analysis. It enables formal incorporation of prior beliefs, which are updated by the data to give a so-called posterior distribution for probabilistic inference using all relevant information. For the way of learning under uncertainty as evidence accumulates, there is a big role for Bayesian statistics to play in making clinical trials more efficient. Specifically, its strengths have been primarily seen in the incorporation of (1) historical data from previous studies investigating the same or a similar treatment, and (2) interim analyses for data monitoring or modifications to an ongoing clinical trial. This talk will revisit several case studies that have implemented Bayesian methods for the trial design and analysis, highlighting the added efficiency achieved. Basic steps for establishing a new Bayesian design, as well as performing the analysis, will be covered. Perspectives will also be given on the future development of Bayesian methods with application to small population clinical trials.