A series of dynamic progressive collapse tests using the uniformly distributed load (UDL) was conducted, to analyze the robustness of three precast concrete (PC) beam–column assemblies using dry connections of top-and-seat angles (TSA-D), strengthened top-and-seat angles (STSA-D), and high-ductility reinforcement (DSTSA-D), under a middle column removal scenario. Finite element (FE) models were also developed to accurately simulate the collapse process. Based on the tests and FE models, comparative studies were conducted on dynamic and static collapse responses of the PC assemblies with identical configuration, loading regime, and boundary conditions. The results showed that: the horizontal reaction force developments were similar under dynamic and static collapse scenarios; under dynamic collapse scenarios, the initial stiffness of the PC assemblies increased due to the strain rate effect, while the peak vertical reaction forces under the compressive arch action (CAA) were smaller than those under static collapse scenarios, attributed to dynamic damage; under both dynamic and static collapse scenarios under UDL, the beam deformed in a downward convex curved shape, with local material deformation becoming more concentrated under dynamic collapse scenarios. An energy-based method for calculating dynamic collapse resistance was evaluated and revised: 1) the traditional method, which converts static responses from static analysis into dynamic responses was found to underestimate the resistance under small deformations due to neglecting the strain rate effect, and to overestimate ultimate resistance by ignoring dynamic damage; 2) by using dynamic vertical reaction forces instead, a more accurate prediction of the dynamic collapse resistance was achieved with a single dynamic collapse analysis. Additionally, dynamic amplification factors for the PC assemblies were calculated based on FE models and static test results providing basic knowledge in whole PC structural level research.