In view of faults of the current 3D printing technology such as the long processing time and low production efficiency, a path optimization algorithm of 3D printing based on fused deposition modeling was developed. Because the slice contour of some printing parts was comprised of many closed curves, the starting point of each closed curve was determined by the nearest neighbor method and the 3D printing contour path planning was transformed into the trareling salesman problem. A contour path planning algorithm based on ant colony algorithm was developed in order to plan printing sequence of the contour paths reasonably. Parallel scanning method was adopted for filling the slice cross-section. Influence on printing efficiency and quality for different scanning angles was analyzed and optimal scanning angle was selected. The scanning area was divided into different areas and a region merging algorithm based on four-point method was developed. The printing sequence of the different areas was optimized by the nearest neighbor algorithm, which improved the scanning efficiency and molding quality. The experiment of printing parts was done by means of the path optimization algorithm. Compared with the traditional parallel scanning method, single layer contour path scanning lengths of the three parts were decreased by 19.5%, 12.5% and 10.7%，the printing time was decreased by 12.6%, 11.6% and 8.9% and the dimensional accuracy and surface quality were improved. The result showed that the path optimization algorithm was robust and effective and its efficiency was high.