Hospital inpatients and elderly healthcare facility residents are often at risk of falling in the toilet room. Quickly finding fallen persons makes it possible to prompt rescue them. Thermal imaging sensors can provide a method for fall detection in toilet rooms that is superior to other detection techniques. The present study used a thermal imaging sensor to differentiate between normal toilet room activity and simulated falling activity by healthy subjects, after which the discrimination rate between those activities was determined. The results showed that the thermal imaging sensor could detect various falls in the toilet room with a discriminant ratio of 97.8%. Also, while the room temperature needs to be 31°C or less, falls can be judged in less than one second, and the subject's privacy can be protected. ©2009 IEEE.