The carrier status of 39 at-risk individuals in 6 multiple endocrine neoplasia 2A families was determined using a DNA based test. We were able to calculate a virtual diagnosis (probability greater than 95%) for 77% of the individuals and a probable diagnosis (probability greater than 90%) for 90% of the individuals. This study points out some of the problems of specific pedigree structures that can affect the risk calculation. This study further shows that no single test based on either biochemistry, pathology, or genetics can consistently and unambiguously produce a presymptomatic diagnosis. We also describe two specific examples where DNA testing has helped to resolve clinical uncertainties in at-risk individuals.