Hi, I posted suggestions for improving the prim minimum spanning tree implementation a while ago[1]. This is a 'ping' mail to see if its possible to get further comments on how to progress forward. Benchmarks summarized: new prim_minimum_spanning_tree 47.585s current prim 3m26.38s (338% slower) kruskal_minimum_spanning_tree 40m57s (5063% slower) 1. http://lists.boost.org/Archives/boost/2006/08/109553.php -- Magnus Ekdahl