In situ sensors that are used to detect the pavement response play an important role in the health monitoring of asphalt pavement. However, embedded sensors have both the positive benefits of pavement condition monitoring and the negative effects of structural-performance degradation, especially for sensors with improper external shapes or profiles. The stress and strain field variations and sensor interaction with damage should be included in the structural optimization problem. In this article, the effect of sensor embedding on the stress and strain field variations was analyzed through numerical studies on an asphalt concrete specimen under three-point bending configuration. Three-dimensional finite element models of sensors with different shapes (I-shaped sensor and bamboo-shaped sensor) were developed using ABAQUS to compare the dynamic response under cyclic loading. The analysis results indicated that the embedment of a bamboo-shaped sensor caused greater mechanical response and enhanced stress concentration at the position where the sensor was embedded. In particular, the embedment of two types of sensors had little influence on the stress and strain field inside the asphalt concrete beam within the depth range of 0–60 mm, but different degrees of influence were produced within the depth range of 60–100 mm (tension zone). The results of axial stress and strain indicated that the deformation of I-shaped sensor was more uniform and coordinated with that of asphalt concrete. Afterward, the excellent compatibility and deformation coordination between the I-shaped sensor and asphalt concrete was confirmed by three-point bending tests and further applied to real pavement structure to monitor traffic loads. It can be concluded that the material and structural compatibility of I-shaped sensor are superior to that of bamboo-shaped sensor as a pavement sensor, and the dynamic response obtained by the I-shaped sensor is closer to the true value. The study results provided an optimization structure for design reference of embedded sensors for pavement structural health monitoring.