Abstarct: One of the common issues in the cast-in-situ reinforced concrete structures is creating a construction joint (cold joint) caused by an interruption or delay in the concreting operations. The cold joint is a weakness or defect in the concrete, which results in the non-integrity of the concrete. For this reason, the performance of concrete elements with the cold-joint is under the influence of that behavior. The seismic design procedure for in-situ construction generally considers that the connection of beam and column that frxames into the joint is monolithic in nature. But in actual construction, it is not possible to cast columns of the multi-story frxame in one go and therefore, a cold joint is inevitable in all the upper story columns immediately above the lower story slab. In this research, firstly, cold joint behavior is modeled. The model of concrete damage plasticity used for the modeling the concrete behavior and the surface-baxsed cohesive behavior with the traction-separation response used for the modeling the cold-joint. The three-point bending beam specimens with the same compressive strengths of concrete on both sides of the cold-joint have been used to verify the opening mode behavior of the cold joint from the experimental results. Three different sizes of the beam were considered to ensure the validation of opening mode behavior for the cold joint. So, the push-off test specimens have been used to verify the shear-friction behavior of the cold joint from the experimental results. Three same specimens with same compressive strengths of concrete on both sides of the cold-joint and the different number of steel connectors crossing the interface surface of the push-off specimens were considered to ensure the validation of shear-friction behavior for the cold joint. Then, a single-story single-bay reinforced concrete frxame is modeled. After ensuring the validity of the numerical model of the cold joint and frxame, a reinforced concrete frxame containing a cold joint is modeled on its columns at the below of the beam and the top of the foundation. Subsequently, in order to investigate seismic behavior, an In-plane monotonic loading, stroke-controlled pushover tests were performed once on a frxame containing a cold joint and once again on the same frxame but without a cold joint. The frxame was also modeled in two situations with two and three spans in both conditions of monolithic and with a cold joint. From the result, prior to the yield point, there was no significant difference between the ultimate load of the monolithic frxame and frxame with cold joints. A significant effect on the frxame behavior was achieved in the monolithic frxame and frxame with cold joints in their ultimate displacement so that the ultimate displacement in the cold-joint state for one, two and three was reduced by about 17, 31 and 31 percent, respectively, compared to the monolithic one.