Natural language processing, particularly in the form of statistical models, is being used in more speech applications. This session introduces natural language processing and its role in speech applications. The key ideas are the following: what natural language is, the statistical language model (SLM) approach to natural language processing, when and how to use natural language processing techniques in an application, how to combine natural language processing techniques with grammars and directed dialogues to achieve optimal application performance, commercially available natural language tools, a brief discussion of research areas and newer technology such as the technology used in the IBM Watson Jeopardy system. This tutorial is aimed at an audience with a general technical background. Experience developing speech applications would be helpful. Attendees can experiment with a simple open-source system that illustrates the key concepts of SLMs.