Speech recognition is one of the important application areas of artificial intelligence. Speech recognition is the process of converting the acoustic signals into corresponding text transcription. Achievement of the high speech recognition accuracy is the critical factor. It is very complex task to identify the speech spoken in different rates. Duration of the speech is one of the most important features used to identify the speech at particular rate. The present paper focuses on the prediction of the speech using duration parameter and classifies them in to slow, normal and fast.