Historical data and real world data (RWD) are increasingly available through big data initiatives. Researchers are enthusiastic about how to use these data to better inform clinical trial designs and drive decision making. Selection bias, missing data, and heterogeneity in the historical data could lead to undesirable outcome in both design and conclusion of a trial utilizing such information. In this presentation, we summarize our experience by showing case studies from rare disease and oncology and further discuss the role and impact of historical data in clinical trials. We also discuss some statistical methods including Bayesian, frequentist, predictive modeling via machine learning approaches to help 1) inform clinical trial design; 2) identify prognostic factors and the right population to treat; 3) assess feasibility of study; and 4) increase probability of success or early termination. Ultimately, we hope to be able to better utilize historical data and RWD to expedite drug development process via sound statistical methods to help relieve patient burden and improve health.