Research on neuropsychiatric disorders has produced a number of very important findings in the last few decades. However, several problems continue to hinder research in this area. One problem area has been the appropriate classification of disease status for probands and extended family members in linkage studies. In this article, we examine rates of misclassification in a 12-year follow-up study of children previously diagnosed with Tourette syndrome. At the 12-year follow-up, we found a 5 to 12% rate of misclassification of previously diagnosed cases. We present a model of a linkage study with three classification steps. The model demonstrates that an error rate of 5% would result in misclassification of 20% of true cases by step three. Adding additional steps to improve diagnostic accuracy may increase rather than decrease classification error.