In this paper, a method to predict the healing effect of induction heating asphalt concrete at vertical depths was proposed and verified. At first, AH-70 asphalt concrete and SBS modified asphalt concrete of AC-13 gradation were designed and prepared. R-DT (Relationship of Depth and Temperature) was quantified through induction heating test. R-HT (Relationship of Healing rate and Temperature) was quantified through fracture-healing–fracture test and X-ray micro computed tomography test with oven heating. Then, linear regression was used to verify the rationality of R-HT compared with the pre-relationship in previous studies. Finally, R-DH (Relationship of Depth and Healing rate) was predicted by fitting approach. It was found that R-DT and R-HT presented significant linearity, the existence of which was independent from the type of asphalt materials. While, more sensitive of SBS modified asphalt concrete showed that the response to gradient characteristics of R-DT was related to the higher overall temperature. Additionally, R Square-values = 0.963 of linear regression proved the oven heating method was suitable for the quantification and prediction of gradient characteristics. Furthermore, the prediction of R-DH was obtained successfully, which established the continuous distribution of the healing effect at vertical depths and revealed the linear relationship of the continuous distribution.