Go 1.7 introduces the Run method to tests and benchmarks in the testing package. These innocently looking methods enable a wide range of functionality as often seen in other testing frameworks, but staying true to Go's design principles. This talk will cover how to use the Run methods for setup and teardown, dynamic tests, exerting fine control over parallelism, programmatically creating benchmarks, and selecting subtests and sub-benchmarks on the command line, among other things.