To explore the influence of scope of practice and patient outcomes on error reporting, 13 nurses were interviewed after they reviewed four "error" scenarios ranging in both scope of practice and seriousness of outcome. Of 52 theoretical incidents, only 30 were identified as errors. The nurses indicated they would formally report errors for only eight of the incidents. For another 10 incidents, the nurses would have reported using an informal reporting system only. Qualitative analysis of the interviews revealed that perceived scope of practice influenced reporting preferences, and seriousness of outcome was only a secondary consideration. Selective error reporting and the reasons for selective reporting have negative implications for patient safety.