Preventing heat-related illnesses, which are on the rise due to intense heat, constitutes a critical social issue. The monitoring of core body temperature presents an efficient measure, as heat-related illnesses disrupt the balance of thermoregulatory functions. Directly measuring core body temperature is typically highly invasive and impractical for use during physical activity. The heat flux method, which represents one of the non-invasive core body temperature estimation methods utilizing a skin-attachable heat flow sensor, has been reported as an appropriate approach for active measurements. One notable practical challenge revolves around the occurrence of significant measurement errors due to fluctuations in ambient air temperature and airflow. Here we propose a novel core body temperature sensor grounded in this method, designed to be resilient to variations in the surrounding environment. We substantiate this through both theoretical and experimental verification. Utilizing numerical analysis, we investigated the impact of the proposed sensor structure on estimation. Our theoretical analysis demonstrates the potential to estimate core body temperature more accurately than is achievable with the conventional method, particularly when the ambient temperature and airflow are variable. Furthermore, in line with the outcomes of our numerical analysis, we conducted experimental demonstrations showing the accurate estimation of core body temperature, even under changing airflow conditions. These results strongly validate the efficacy of our proposed new core body temperature sensor, which exhibits resilience to changes in the surrounding environment, both in theory and practice. This outcome is poised to significantly enlarge the potential application scope of non-invasive core body temperature sensors.