This paper presents a High-altitude infrared thermal dataset, HIT-UAV, for object detection applications on Unmanned Aerial Vehicles (UAVs). HIT-UAV contains 2898 infrared thermal images extracted from 43470 frames. These images are collected by UAV from schools, parking lots, roads, playgrounds, etc. HIT-UAV provides different flight data for each place, including flight altitude (from 60 to 130 meters), camera perspective (from 30 to 90 degrees), date, and daylight intensity. For each image, the HIT-UAV manual annotates object instances with two types of the bounding box (oriented and standard) to address the challenge that object instances have a significant overlap in aerial images. To the best of our knowledge, HIT-UAV is the first publicly available high-altitude infrared thermal UAV dataset for persons and vehicles detection. Moreover, we trained and evaluated the benchmark detection algorithms (YOLOv4 and YOLOv4-tiny) on HIT-UAV. Compared to the visual light dataset, the detection algorithms have excellent performance on HIT-UAV because the infrared thermal images do not contain a significant quantity of irrelevant information with detection objects. This indicates that infrared thermal datasets can significantly promote the development of object detection applications. We hope HIT-UAV contributes to UAV applications such as traffic surveillance and city monitoring at night. The dataset is available at https://github.com/suojiashun/HIT-UAV-Infrared-Thermal-Dataset.