A system and method for providing multi-modal messaging for a wireless communication system. A wireless terminal is connected to a wireless access network, wherein the wireless terminal is capable of generating a voice request. A multi-modal message server is connected to the wireless access network for receiving the voice request generated by a user of the remote terminal. A voice recognition application is operable to identify words contained in the voice request. A response generation application uses the identified words to generate a response to the voice request. A unified messaging application is operable to interpret the response and generate a first and second response that is based on the contents of the response generated by the response generation application. The first and second responses are then transmitted to the wireless terminal in different communication formats.