Due to the temperature sensitivity of asphalt mixtures, the synergistic deformation between embedded strain sensors and asphalt mixtures may be poor at certain temperatures, resulting in less accurate strain measurements. Therefore, the main purpose of this article is to consider the synergistic deformation between asphalt mixtures and embedded sensors and to provide guidance for the reasonable design of embedded strain sensors for asphalt pavements. Firstly, the finite element analysis and laboratory tests were used as the main approaches to analyze the main factors affecting the synergistic deformation between the embedded strain sensor and the asphalt mixture. Then, critical design requirements and optimization initiatives for embedded strain sensors dedicated to asphalt pavements were proposed. Finally, the optimal embedded strain sensors were further developed and the proposed design requirements were validated. The results show that the output of the sensor can be consistent with the deformation state of the asphalt mixture only if the equivalent modulus of the embedded strain sensor is the same as the modulus of the asphalt mixture. However, asphalt mixture modulus is susceptible to temperature, and it is difficult to keep the asphalt mixture modulus consistent with the sensor equivalent modulus at different ambient temperatures. Therefore, embedded strain sensors with low equivalent modulus and no encapsulated reinforcement are recommended to monitor the strain of asphalt pavement over a wide range of temperatures. The corresponding optimal embedded strain sensor was developed using low modulus polyimide as the elastic strain beam and silicone rubber as the flexible wrapping layer. The optimal embedded strain sensor has a maximum measurement error of only 4.5% over a wide temperature range. Overall, this article provides a reference for the accurate measurement of strain sensors for asphalt pavement.