Non-uniform radiant floor heating is increasingly explored for improving localized and personalized thermal comfort. Accurate mean radiant temperature (Tr) measurement is crucial due to its separation from air temperature (Ta) unlike convective systems. This study evaluates the accuracy of Tr obtained using the globe thermometer versus a novel mini.RES cube sensor and examines the differences between Ta and Tr in non-uniform heating environments. Experiments were conducted in a 62 m3 climatic chamber in Fribourg, Switzerland, with four cases: one without heating, two with non-uniform heating to simulate radiant asymmetry, and one with uniform heating for comparison. Additionally, an uncertainty analysis was performed to evaluate measurement precision for the Tr measurements using the globe thermometer. The results indicate that the globe thermometer method can introduce errors of up to 4% in determining the Tr. Moreover, the uncertainty values for the Tr values were found between 1.60 ℃ and 2.31 ℃. In cases with non-uniform heating available, the Tr was found to vary by more than 2 ℃ than the Ta, highlighting the need to consider Tr separately when assessing thermal comfort. These findings emphasize the error of obtaining the Tr with globe thermometer and the difference between Tr and Ta in non-uniform heating scenarios.