Background: Alerts, a key feature of electronic health record systems, intend to improve patient safety by providing timely information at the point of care. However, many electronic health record systems generate excessive alerts that are not immediately clinically relevant and that contribute to alert fatigue. Despite growing recognition of alert fatigue as a safety concern, clinicians’ experiences of alert fatigue and the broader system-level factors that contribute to it being experienced are not well understood. Objective: This study aims to use a human factors approach to (1) comprehensively explore how alert fatigue is experienced by junior doctors; (2) identify factors that contribute to experiences of alert fatigue; (3) identify perceived impacts of alert fatigue on employees, organizations, and patients; and (4) identify strategies to reduce alert fatigue in practice. Methods: Semistructured interviews were conducted with junior doctors working in hospitals across Australia. Data were thematically analyzed using a hybrid inductive and deductive approach, informed by the Systems Engineering Initiative for Patient Safety and an information processing model. Results: A total of 20 junior doctors were interviewed. Alert fatigue was described to occur at different stages of information processing, including when alerts were not detected, were superficially processed using mental shortcuts, or required excessive cognitive effort to interpret. When alerts were not detected or thoroughly processed, participants more often perceived impacts on patient safety and care quality due to the potential to miss important information. Further, when alerts required excessive cognitive effort, participants frequently reported interruptions, frustration, and time and effort loss as impacts. Factors influencing experiences of alert fatigue were identified in all Systems Engineering Initiative for Patient Safety work system domains, including those related to people, tasks, the environment, tools and technologies, and the organization. Key contributors included the design and clinical relevance of alerts, institutional norms and expectations, and information overload from system alerts as well as other alerts and tasks. Alert fatigue was also described to be experienced differently depending on provider characteristics, such as experiences with and knowledge of alerts, mood, and personality, and organizational factors, including culture, shift type, and time of day. Conclusions: Alert fatigue is not a binary concept but is instead experienced on a continuum and influenced by interacting individual, technical, and contextual factors. Future research should incorporate clinician self-reports to evaluate experiences of alert fatigue in addition to objective measures. Addressing alert fatigue requires tailored interventions that target its different causes and outcomes. These could include technical and design improvements, changes to organizational practices, and individual customization to reduce experiences of fatigue and accommodate differences in clinicians’ needs.