The study was to evaluate the air-conditioning of sleep algorithm. The algorithm was developed through the analysis of brain waves and MST, the experiments using air conditioner was performed in a apartment bedroom. Five female subjects were participated for the experiment. Eight hours of data collection a day was performed under different algorithm, case A, case B and case C. Physiological signals, EEG, ECG, EOG, and EMG, were obtained using polygraph and converted into digital signal. Then, subjects were asked to answer the questionnaire about their thermal sensation after experiment in bedroom. Sleep stages were classified, then TST, Sleep latency and Sleep efficiency were calculated for the three different air conditioner algorithm. As results, TST, Sleep efficiency, questionnaire showed the higher values for Case B algorism than that for other algorism. On the other hand, SWS latency was lower than for other conditions. Therefore, it was concluded that Case B of the algorithm was the best for comfortable and deep sleep.