WSR Macros combines the ease of Windows Speech Recognition with the power of an XML-based scripting language. It makes it easy to create speech interfaces to applications and perform otherwise tedious tasks with simple spoken commands
Check out Rob’s Blog for more details
Also, Will DePalo (a Microsoft MVP) posted two videos on MSDN that shows how to program against the built in speech APIs. Check it out!