This presentation describes two examples in which sponsors performed randomization using an Interactive Voice Response System (IVRS) so that treatment codes for individual patients were no longer available at the sites for inspection. In addition, copies of laboratory source documents that comprise a critical component of primary endpoints may also not be available at the sites. These two examples demonstrate that in such a randomization set-up data manipulation and fraud are difficult to detect by FDA inspectors. However this should not deter a biostatistician and other members of the review team from investigating data that appear irregular or too good to be true.