The main objective of this study was to develop a simulation program to determine the sample size for a clinical study to confirm a genetic-disease association observed in a retrospective exploratory study. The effect of misclassification of a binary response variable on the power is also investigated. A general expression for the magnitude of the decrease in statistical power due to misclassification is obtained based on the Pitman asymptotic relative efficiency. The simulation program presents an estimate of the exact power when misclassification exists. Running the program several times under different settings of parameters, it revealed that the effect of even low misclassification rates is serious. Response misclassification should be taken into consideration when determining the sample size. The program can be used on the Internet.