Abstract: The PV hosting capacity(PVHC) of distribution networks is constrained by the line carrying capacity. However, the traditional assumption that the carrying capacity of overhead lines is static may lead to an inaccurate assessment of PVHC of distribution networks. In the light of this, a PVHC assessment model for distribution networks is proposed, taking into account the dynamic thermal rating(DTR). Firstly, a voltage sensitivity matrix is introduced to quantify the impact of PV access on the voltage distribution of distribution networks, and a distributed PV site-selecting method is proposed based on the traversal idea. Subsequently, in consideration of the influence of meteorological factors on the carrying capacity of overhead lines, a PVHC assessment model for distribution networks taking DTR into account is established through meteorological data-driven statistical analysis. Given the non-convexity of the constructed model, the second-order cone relaxation technique is adopted to transform it into an easy-to-solve mixed-integer second-order cone programming problem. Finally, the proposed model is simulated on the IEEE 33-bus system using actual data. The results indicate that the proposed method can effectively improve the PVHC of distribution networks. Additionally, it also avoids the overload risk of the line currents exceeding the actual line carrying capacity, which occurs as the DTR becomes lower than the STR under harsh meteorological conditions. In summary, the proposed method holds significant application and engineering value.