Clinical alarms have a deceptively simple purpose, which is to notify caregivers when a patient or a device needs their attention. This simple concept has been proven to be challenging as the number of available alarms has grown and been poorly integrated. When an anticipated notification is not received or an actual notification is not acted upon in a timely manner, patient harm can occur. In this regard, false alarms have been proven to be highly detrimental to the effective use of clinical alarms to enhance patient care. Equally problematical is the issue of false reliance in which a clinician's vigilance is degraded by the expectation that if anything bad happens, the system will notify him or her. Similarly, alarms have also been part of staff downsizing and shifting to lower expertise, wherein it is believed that the alarms are an appropriate substitute. Human factor issues associated with setting, observing, and responding to alarms have also been proven to be inadequately addressed. This article presents a fault tree analysis of the patient harm-related clinical alarms failures. This analysis can be used to understand, debate, and educate.