This paper introduces genetic algorithms for the level permutation pro
blem (LPP). The problem is to minimize the number of edge crossings in
a bipartite graph when the order of vertices in one of the two vertex
subsets is fixed. We show that genetic algorithms outperform the prev
iously known heuristics especially when applied to low density graphs.
Values for various parameters of genetic LPP algorithms are tested.