With highly active targeted therapies becoming more commonplace in recent years, strategies for development become increasingly important in terms of time-to-market, clinical trial size and the breadth of patient population that may benefit from a drug. We discuss strategies and corresponding statistical tools that have been applied. Early single-arm trials followed quickly by randomized trials is a bedrock approach for many indications, but not without pitfalls. The possibility of doing trials that are target-based rather than histology-based as well as other novel approaches are considered. Another challenge is developing one or more biomarkers at the same time that a drug is being developed.