Carbon monoxide (CO) is a dangerous gas with no odor, color, or taste and can easily mix with other air around it. This study aims to design and determine the results of measuring CO levels using a drone integrated with an MQ-7 gas sensor and an IoT-based ESP32 nodeMCU by manipulating the time and height of measuring CO levels. The method used is data acquisition in the form of a data retrieval process from the MQ-7 sensor programmed with the ESP32 nodeMCU via the ThingSpeak web. The results of measurements of CO levels in Ketintang, Surabaya air quality in the medium category at 07.00-11.00 WIB and good at 15.00 WIB. Measurement at 07.00 WIB CO levels are higher than at 11.00 WIB and 15.00 WIB. This can occur because the use of motorized vehicles and the process of burning waste in households is denser. Most of these activities are carried out close to the ground so that the higher the altitude in measuring CO levels, the lower the CO levels detected. Other factors that affect CO levels are temperature, wind speed, and weather when taking measurements. Based on the results of the study, it can be concluded that hours and altitude affect the CO levels produced.