A method of speech to DTMF generation involving ASR-enabled and DTMF-controlled communications systems. The ASR-enabled system is used to recognize speech received from the DTMF-controlled telecommunications system using sampling rate independent speech recognition. It then identifies a speech segment contained in the speech received from the DTMF-controlled system that corresponds with at least one keyword associated with user-defined data. Then, the ASR-enabled system transmits at least one DTMF signal to the DTMF-controlled system in response to the identified speech segment. This allows a user of an ASR-enabled system such as a vehicle telematics unit to at least partially automate access to the DTMF-controlled system using the telematics unit, so that voice mailbox numbers, passwords, and the like normally entered via a telephone keypad can be automatically sent to the DTMF-controlled system from the telematics unit without having to be manually input each time by the user.