Abstract: This study proposed a constrained and adaptive regularized 1D scheme for the vertical component inversion of semi-airborne transient electromagnetics. Based on the Occam inversion, the CMD adaptive regulation scheme was used to calculate the Lagrange multiplier. Meanwhile, the feasible descent direction method with natural boundary conditions and model correction was introduced to constrain the inversion process. This combined algorithm can both improve the computational efficiency of the inversion process and guarantee the stability and reliability of inversion results. The results of layered models show that this algorithm can obtain ideal inversion results with a short iteration and computation time. Owing to the shielding effect of conductive layers and the insensitivity to resistive layers of transient electromagnetics, the inversion results of the HK model can only get the average resistivity of underlying layers. The results of the HK model verify the adaptability and effectiveness of this algorithm proposed in this study for complex geoelectric structures. A satisfactory inversion result was achieved from noise-included signals through no more than ten iterations. This algorithm exhibited good stability and convergence in numerical simulation, thus verifying that it is a feasible and effective method for interpretations of multi-source semi-airborne transient electromagnetic data.