Effective tsunami risk reduction requires an understanding of how at-risk populations are specifically vulnerable to tsunami threats. Vulnerability assessments primarily have been based on single hazard zones, even though a coastal community may be threatened by multiple tsunami sources that vary locally in terms of inundation extents and wave arrival times. We use the Alaskan coastal communities of Cordova, Kodiak, Seward, Valdez, and Whittier (USA), as a case study to explore population vulnerability to multiple tsunami threats. We use anisotropic pedestrian evacuation models to assess variations in population exposure as a function of travel time out of hazard zones associated with tectonic and landslide-related tsunamis (based on scenarios similar to the 1964 M w9.2 Good Friday earthquake and tsunami disaster). Results demonstrate that there are thousands of residents, employees, and business customers in tsunami hazard zones associated with tectonically generated waves, but that at-risk individuals will likely have sufficient time to evacuate to high ground before waves are estimated to arrive 30–60 min after generation. Tsunami hazard zones associated with submarine landslides initiated by a subduction zone earthquake are smaller and contain fewer people, but many at-risk individuals may not have enough time to evacuate as waves are estimated to arrive in 1–2 min and evacuations may need to occur during earthquake ground shaking. For all hazard zones, employees and customers at businesses far outnumber residents at their homes and evacuation travel times are highest on docks and along waterfronts. Results suggest that population vulnerability studies related to tsunami hazards should recognize non-residential populations and differences in wave arrival times if emergency managers are to develop realistic preparedness and outreach efforts.