Statistical analyses of biomarker studies involve new challenges. This session will give an overview of the multiplicity issues involved. We will first review fundamentally the reason for controlling (Type I) error rates, in a regulated industry. Then we review why controlling Tukey’s (1953) familywise error rate (FWER) has been the practice in clinical trials. Finally, we will consider alternative error rates, more liberal than FWER, that may be useful in biomarker studies. These include Tukey’s per family error rate, Benjamini and Hochberg's false discovery rate (FDR), Efron’s local Fdr, and generalized Familwise error rate.