Stanley Young National Institute of Statistical Sciences The discovery of bioactive molecules is a complex, time consuming and expensive activity. Many millions of molecules are available for testing. Automation is working on one side of the problem. Statistical strategies are working on the other side. Work hard or work smart? Our idea at GlaxoSmithKline was to partner with the North Carolina State University Statistics Department using graduate students and faculty. Dr. Jacqueline Hughes-Oliver became our NCSU collaborator. Thesis work was done by multiple students. Problems, data and intellectual involvement came from GSK. Hard work and intellectual guidance came from Dr. Hughes-Oliver. A very productive NIH grant produced state of the art analysis software. Pooling strategies were developed. A new recursive partitioning method was invented. Ultimately, all the major drug companies took up the challenge of using statistical methods to work smarter on drug discovery.