The pharmaceutical industry depends on accurate and reproducible data science for both preclinical and clinical analysis. Unfortunately, often an analysis cannot be reproduced and therefore its computational methodology and merit are unknown. Often, the data, code, or description of computational methods is not maintained. In order to implement good practices of reproducible computational research, the leadership of the company must invest time and resources into planning, training, ensuring adoption of common practices and tools, implementing documentation systems, encouraging discipline on the individual and group level, creating incentives, and requiring accountability. At Eisai, we have developed a working system for reproducible computational research that is enabled by leadership, technology, and culture. With regards to technology, we primarily use Rmarkdown and R Notebooks on an Rstudio server used by all our analysts. The Rstudio server is maintained by an administrator who installs packages for all users, creating a common package environment that ensures that code can be rerun in the future. Data and code are stored in a shared network drive and version control is accomplished by using Git. A wiki that is editable by all analysts is used to organize all analyses (tracked with unique analysis IDs) and provides links to code and results. With regards to culture, the leadership has promoted the values of quality and reproducibility. When yearly objectives are set, the performance criteria includes the creation of analysis documents (e.g. Rmarkdown reports), use of version control, and organization of data on shared network drives. Setting aside time for wiki documentation in the midst of high demands from project teams is helped by having periodic “documentation day” parties. To verify reproducibility, we have implemented “witnessing” once the analysis is finished, it is reviewed by an independent team member who officially signs off on the work, stating that the reproducibility criteria have been met. Our success in implementing reproducible computational research can serve as a model for other companies to use. Here we have provided a model based on leadership, technology, and culture.