Taiwan Sugar Corp.'s Hsianghe Student Dormitory is located in the administrative area of Chiayi County, and it is leased to mostly college students, school staff, and government agency employees. However, in recent years, due to the low birth rate and the construction of a large number of residential buildings for students, the occupancy rate has seriously declined. As an effective solution to the operating difficulties of the Hsianghe Dormitory, this study collected 70 cases in this study to predict the rent of Hsianghe Dormitory by using the BPNN model, expecting to set a reasonable rent to meet consumer demand, increase market competitiveness and enhance the Hsianghe Dormitory occupancy rate. This study used the BPN model and Super PCNeuron 5.0 software to test and analyze the cases. The test models included the training and test model, the validation model and the inference model. The final output value of the inference model is the reasonable rent of Hsianghe Dormitory. The data of cases were collected in seven counties and cities including Taipei City, New Taipei City, Hsinchu City, Taichung City, Chiayi County/City, Tainan City, and Kaohsiung City. Sixty cases were collected for the training and testing model, 10 cases were collected for the verification model, and the final inference model used the Hsianghe Dormitory as the inference subject. Regarding the reference variables, this study summarized the main impact factors of renting as described in literature including location, daylighting, facilities, service quality, security as the benchmarks for assessment. In the inference model, regarding the current conditions of the dormitory of Hsianghe Dormitory, this study rated the “location” in Chiayi County as 1 point, the balcony and French window for bright “daylighting” as 5 points, the “facilities” including elevator, TV, refrigerator, air-conditioning, parking space and monitored door access system as 5 points, the “service quality” including service personnel on three shifts as 5 points, “security” in the charge of specifi c personnel as 5 points. After inputting the data for testing the proposed model, the result from the model was more reasonable as compared with the prevailing rents of the area.