[en] In a probabilistic risk assessment (PRA), estimates of the probabilities of occurrence of a wide variety of basic events, such as equipment failures and human errors, are combined to provide estimates of other parameters of interest such as core-melt frequency. Identification of important basic events in a PRA can help identify weak spots in the power plant and lead to possible improvements. Thus, to this end, a variety of importance measures have been proposed and used to help interpret PRAs. These measures are described here and the results of applying these measures to PRAs of the Millstone(3) and GrandGulf(4) plants (two boiling water reactors, BWRs) are presented. The Grand Gulf Reactor Safety Study methodology Applications Program (RSSMAP) and Millstone Interim Reliability Evaluation Program (IREP) PRAs were selected due to the quality of the core-melt frequency equations presented in the PRAs and because the plant chosen for the RMIEP PRA, La Salle, is a BWR. The results from this analysis will help guide data collection efforts for the LaSalle PRA