Intelligent home service systems consist of ubiquitous sensors, a home network, and a context-aware computing system that together collect residential environment information and provide intelligent services such as controlling the environment or lighting. Determining a resident's location in the smart home or smart office is a key to such a system. This correspondence presents an enhanced location-recognition algorithm using a Bayesian classifier for the pyroelectric infrared sensor-based indoor location-aware system that is a nonterminal-based location-aware system proposed in a previous paper. This correspondence compares the conventional and enhanced location-recognition algorithms and their performance. The feasibility of the system is evaluated experimentally on a test bed.