Regulatory-grade patient-level historical clinical trial data can greatly improve our ability to design efficient clinical trials and understand key characteristics of disease indications. We describe a method for constructing and utilizing a well-characterized historical control arm, which we call a Synthetic Control Arm (SCA), using data from patients in earlier clinical trials. These patients are drawn from clinical trials hosted by Medidata Solutions for pharmaceutical companies, CROs, and research organizations around the globe; Medidata has contractual rights to extract and utilize knowledge, suitably aggregated and anonymized, from these trials. We present the conditions under which an SCA can be an effective comparator for a current single-arm trial of interest. We describe the challenges in aggregating and standardizing SCA data. Bayesian statistical methods provide powerful strategies for obtaining information from historical control data. We illustrate the practical advantages and disadvantages of using Bayesian inferential methods in the presence of an SCA.