The final bending angle by a press brake equals the sum of clamp and springback angles. Hence, the on-line measurement of the clamping angle and prediction of springback are key technologies for the precision bending. Recently, a precise angle sensor for the measurement of clamp angle has been developed, and implemented into bending machines for the on-line measurement. If the springback is known a priori, the required clamp angle can be calculated and the clamping position can be decided using a feedback control system with the precise angle sensor. As a result, a high-accuracy V-bend process becomes possible. In this research, a new computational method, based on fundamental bending theory, is proposed for the prediction of springback. However, errors will arise if the material anisotropy or the Bauschinger effect in springback are not accounted for. For the consideration of the Bauschinger effect, a new method is introduced in the fundamental analysis. First, a series of tension tests are conducted for the investigation of the Bauschinger effect for various values of pre-strain and different material types. Based on these results, a new empirical equation is proposed for the description of the Bauschinger effect. The empirical equation is combined with the fundamental bending theory to improve the calculation accuracy of the springback, followed by experimental verification. This method makes it possible to reduce the error margin of springback to less than 0.5 degree with a computing time of about one second and can also be easily implemented on the filed.