While various observation techniques have been developed to measure park use or park-based physical activity, no study has used unmanned aerial vehicles (UAVs) to do so. Thus, this study develops a new observation method that uses UAVs to survey park-based physical activity. This study tests the inter-rater reliability and criterion validity of the UAV-using observation method in comparison to an existing on -the -ground observation tool in five diverse urban parks in Salt Lake City, Utah. With a systematic observation tool, SOPARC (System for Observing Play and Recreation in Communities), this study finds that the UAV observations show a high level of inter -rater reliability (ICC = 0.99 for a total number of users). In addition, compared to the results of on -the ground observations using SOPARC, those of UAV observations demonstrate validity (ICC = 0.98 for a total number of users). Compared to existing methods, the UAV observation tool, covering larger target areas, is suitable for counting park users in a more reliable and efficient way and mapping their use patterns; however, the tool is weaker at collecting detailed user information and surveying under poor weather conditions. Thus, the UAV method could complement existing tools. Finally, this study suggests practical implications of the UAV observation method.