The physical and mechanical behavior of the knitted fabrics is affected by the loop shape. Therefore, the loop geometry is investigated under different conditions in many researches. Most of the proposed models have attempted to provide a realistic approximation of knitted loop, but yarn structure is considered continues in all of them. All the loop models assumed that yarn construction is composed of a monofilament and a uniform cross section. Slippage of any monofilament in the yarn structure rejects continuity of the yarn hypothesis. Even in high amount of twisted spun yarn in which the lateral force on the cross section prevents the fibers movement and also the continuity assumption is rejected for this kind of yarns. Despite the above mentioned reasons, the continuity assumption for the yarn makes the analysis simpler. Modeling the knitted loop shape using a flat multifilament yarn is more complicated than monofilament. Considering this complexity, there is no research work about the multifilament yarn loop model until today. In this work a method for modeling of a multifilament knit loop has been studied. The proposed method consists of three phases including planar post-buckling of multifilament yarn, bending the planar Elastica in 3D space and optimization of yarn structure using genetic algorithm. First phase of the thesis associated to investigating post-buckling behavior of a multifilament yarn in a plane. When a multifilament yarn is buckled in a plane applying two equal and opposite compressive forces and the tangents to the curve at free ends are kept parallel, assuming that the monofilaments can pass through each other, at the top of the knit loop, they have volumetric intersection with each other. These intersections must be eliminated at the second phase. The Post-Buckling behavior of multifilament yarn is a very complex phenomenon because the complicated interactions have been existed between monofilaments. The effects of these loads have been replaced by an out-of-plane bending force at the top of each monofilament. This bending loads cause them to bend inside the plane perpendicular to Elastica plane. At the last phase the amount of such force for each monofilament must be determined using an optimization method. In this case a common evolutionary algorithm called genetic algorithm has been used for the optimization task. Fitness of the optimization determined to be summation of volumetric intersections between monofilaments. To verify the proposed method the yarn has been modeled using finite element method in ABAQUS software. Both modeling methods assume that monofilaments of yarn are incompressible with circular cross section. The model and ABAQUS simulation results are in good agreement, so the proposed model is reasonable.