A proposed nationwide postal questionnaire to Swedish parents who had lost a child due to cancer between 1992 and 1997 was denied approval by the local ethics committee. However, a pilot study to assess the harm and benefit of the questionnaire was approved. 95% of parents found the pilot study valuable; thus, we were allowed to proceed with the main study, which consisted of 129 questions about the child's care and death and five about the parents' perceptions of the study. 423 (99%) parents found the investigation valuable, 285 (68%) were positively affected, and 123 (28%) were negatively affected (10 [2%] of whom, very much). Although the numerical data cannot be directly translated to ethical conclusions, they can provide guidance for future ethical decisions.