More and more clinical trials for new drugs are designed as non-inferiority trials. In order to assess the validity of these trials, however, a number of peculiarities have to be considered. This applies, e.g., to the exact definition of "non-inferiority" as well as to the data evaluation methods.