This paper presents the work done to improve the recognition rate lit an isolated word recognition problem with single utterance training. The negative effect of en ors (due to insufficient training data) in estimated model parameter is compensated by fusing the information obtained fi om HMM evaluation and those generated for the word length and voicing at the beginning and end of the word. A Bayesian Causal Tree structure is developed to accomplish the fusion. The final decision is made on one of the three candidates which ale most likely according to HMM evaluation. The reliability of the HMM ordering is improved by applying variance flooring.