A best-fit curve is required to reconstruct craniofacial fracture to ensure the preciseness of the reconstructed contour. A genetic algorithm (GA) is applied to get the best-fit curve in reconstructing the fractured part. This method provides a fast decision in obtaining suitable sets of control points to interpolate both boundary regions to form a reconstructed part without any try-and-error approach that requires altering the control points several times. The optimised sets of control points with different degrees of continuity are used to develop the reconstructed part using the quintic Bézier curve to generate a smooth curve. The best-fit curvature value of the quintic Bézier curve for each degree of continuity is compared, and the curve with the lowest absolute error of curvature is chosen as the inner and outer parts of the craniofacial fracture reconstruction.