The seed orchards design is the basic work for forest tree breeding. The goal of seed orchards design is to minimize the inbreeding between clones and guarantee the high genetic effects in seed orchards. The data source was genetic distance between the 1st, 1.5th and 2nd seed orchards of Pinussylvestris var. mongolica in Inner Mongolia Honghuaerji Sylvestris National Seed Areas by using SSR molecular marker technique. Considering the complex relationships between parents, the scheme of seed orchards was achieved by improved fruit fly optimization algorithm. The result was compared with other intelligent optimization algorithms and traditional algorithms, and it was showed that the speed and effectiveness of the improved fruit fly optimization were significantly better than those of other algorithms. After executing 200 times respectively, its maximum value was significantly lower than the minimum value of other algorithms. Finally, the improved fruit fly optimization design scheme was compared with the completely random design scheme and the genetic algorithm design scheme. Multiple same clones existed adjacent to each other in the completely random algorithm design, but the genetic algorithm and the improved fruit fly optimization design program did not appear as close neighbors of the same clonal ramets. However, the preferred mating of the improved fruit fly optimization design scheme was superior to the genetic algorithm design scheme. Based on the genetic distance to deal with the complex relationships of the parents, using the intelligent optimization algorithms to design seed orchard can provide reference for the design of highgeneration seed orchards.