There is no magic bullet. I think at the moment some of the programs that provide the best data have, first of all, two levels of prevention: primary prevention for everybody in the schools, for example, and having that there, and then the second level, for people who we identify as being at risk. That starts with children whose parents are already using drugs.
My personal preference would be to start with the people at risk as being a good prevention program, but I don't have a magic prevention program.