In this paper, an algorithm is presented based on using bspline function for optimizing tank cross section. This process minimizes fluid c.g. height and overturning moment and improve rollover threshold of tank vehicles. This algorithm receives tank capacity specifications as inputs and offers fourth order bspline function with 10 control points that has more roll stability, and then optimizes it for different filling conditions. This algorithm is based on the third order bspline function with 8 control points, initially. Therefore, with averaging and optimizing, range of control points is modified and the numbers of control points and degree of bspline function are increased. The results show that, the mutation rate is better to be between 4 and 6%, and the number of individuals in each generation should be at least 40. The algorithm presented in this paper, is a fast and accurate method for optimization of tank cross section in different filling conditions. The Algorithm based on GA maintains simplicity applicable for industries and specially has a rollover threshold of 10% higher than conventional tanks.