Linkage disequilibrium is an important tool both at the end stages of positional cloning studies to map genes of particular interest and in reconstruction of population histories. With advances in molecular biology, complex genetic systems involving multiple highly polymorphic markers are becoming more commonly used to study linkage disequilibrium. These systems contain much more genetic information than simple two marker systems. In this article, we introduce a measure to summarize the overall deviation from random association and propose a permutation-based estimate for this measure. The performance of the proposed estimation procedure is studied through simulations. The methods proposed in this paper are then applied to population data at the dopamine D2 receptor locus (DRD2) and at the homeobox B (HOXB) gene cluster.