These datasets support the conclusions in the journal article entitled "Modeling non-structural strategies to reduce pedestrian evacuation times for mitigating local tsunami threats in Guam" as described in the abstract below: Reducing the potential for loss of life from local tsunamis is challenging for emergency managers given the need for self-protective behavior within brief windows of time for at-risk individuals to evacuate. There has been considerable attention paid to discussing the construction of tsunami vertical-evacuation structures for areas where there is insufficient time to evacuate. This strategy may not be feasible for at-risk populations in island communities for multiple reasons. We examine the influence of three non-structural interventions (reducing departure delays, increasing travel speeds, and managing vegetation to create new paths) that may improve the evacuation potential for at-risk individuals in island communities and use the US territory of Guam as our case study. We model pedestrian travel times out of a modeled inundation zone for a local tsunami generated by a Mw 8.3 earthquake within the Mariana subduction zone. Evacuation-modeling results indicate that reducing departure delays has a larger impact than increasing travel speeds or creating evacuation corridors through heavy brush on reducing the number of at-risk individuals with insufficient time to evacuate. Travel times to safety are shorter than lead times of expected wave arrivals for all at-risk individuals in the tsunami-hazard zone if one assumes all three interventions are implemented.