Wearable sweat sensors, a product of the development of flexible electronics and microfluidic technologies, can continuously and noninvasively monitor abundant biomarkers in human sweat; however, sweat interferences, such as sebum, can reduce sensor reliability and accuracy. Herein, for the first time, the influence of sebum on the potentiometric response of an all-solidstate pH sensor was studied, and the obtained experimental results show that sebum mixed in sweat can decrease the potential response of the sensor and the slope of its calibration curve. A paper-based sandwich-structured pH sensor that can filter the sebum mixed in sweat was proposed based on commonly used oilcontrol sheets. Moreover, the hydrophilic properties, microstructure, and microfluidic performance of the sensor were investigated. The detection performance of the paper-based sandwich-structured pH sensor was comprehensively evaluated in terms of calibration in the presence of sebum and potentiometric response upon the addition of sebum. Furthermore, the antiinterference ability of the sensor was evaluated using different analytes under various deformation conditions. On-body trials were conducted to verify the performance, and their results showed that the proposed sensor can filter over 90% of the sebum in sweat, significantly enhancing sensor reliability and accuracy. Additionally, microfluidic channels could be simply fabricated using a scissor and paper, obviating the need for complex micromachining processes, such as photolithography and laser engraving. Overall, this work illustrates the influence of sebum on the detection performance of traditional potentiometric wearable sensors and paves the way for their development for real-world applications.