Percutaneous needle insertion is widely used in minimally invasive procedures, in which the flexible needle is steered to reach a specific target inside the human body. The targeting error is due to a combination of flexible needle deflection and target displacement in soft tissue and only a very limited number of studies have focused on both two factors. This paper presents a targeting error calculation method which incorporates an energy-based needle deflection model into a soft tissue finite-element (FE) model. The needle insertion process is discretized into several increments on the basis of the quasi-static method. Needle deflection in each step is obtained by the needle-soft tissue interaction model which is applied into the FE model as the displacement input. A 2D-planar FE model is used to model the target displacement by imposing needle distribution forces and needle deflection at different steps on the appointed reference nodes. The soft tissue is modeled as a non-linear hyperelastic material with geometrical non-linearity. Uniaxial tensile strength tests are utilized to determine the soft tissue parameters. Needle targeting experiments are conducted to validate the simulation results. Results show that the proposed method can predict the needle targeting errors while the averaged prediction error stays below 0.4 mm. At last, we conduct different experiments to compensate the obtained targeting error and thus, reaching preferable effects.