The retrospective approach in palliative care research provides valuable insight into death and dying, and the effectiveness of palliative care. The method involves collecting information from proxies (usually significant others) after the patient's death. This exploratory study investigates whether proxies' accounts differ during bereavement, and provides possible explanations for why discrepancies might occur. Thirteen bereaved family members were interviewed, at three to five months and seven to nine months after the patient's death, about the patient's pain, anxiety, and depression, using semi-structured interviews and the symptom rating scale from the Views of Informal Carers-Evaluation of Services (VOICES) interview. Analysis of VOICES ratings over time indicated consistency for anxiety, while pain and depression ratings were variable and, in many cases, less severe and less frequent with the passage of time. Qualitative analysis of proxies' interview transcripts revealed a number of categories and themes that could be explained within the psychological and palliative care literature. The findings suggest that timing is an important consideration when gathering information from proxies retrospectively.