Compared to the isotropic dielectric elastomer (DE), uniaxial fiber reinforced DEs have larger deformation capabilities in specific directions, showing great application prospects. However, there are still significant challenges in achieving self-sensing without introducing additional sensing components. Inspired by the control method of biological muscles, in this article, we used carbon nanotube (CNT) fibers as reinforcing fibers that resistance changes are highly sensitive to tensile strain and exhibit a linear relationship within a small strain range. Based on the developed uniaxial reinforced strain behavior model of DEs, a shear lag model was adopted to introduce an interface transition layer between the DEs and CNT fiber to consider the strain difference caused by the difference in elastic modulus. The relationship between CNT fiber strain and DE strain was established, and the influence of CNT fiber mechanical properties on self-sensing performance was explored. The verification results under 0.05 Hz, 20 kV/mm signals show that the model has 4 % error, providing new ideas for the self-sensing function of uniaxial fiber reinforced DEs.