The aim of this was to derive population-based norms for women completing the EORTC QLQ-C30 version 1 which is designed for use with patients who have cancer. The study was conducted using two different questionnaires: one designed for use in female patients with breast cancer, the other for those with gynaecological cancers, but both including the EORTC QLQ-C30. The women were drawn from the Danish Central Population Register without knowledge of their health status and divided at random between the two questionnaires. All procedures for collecting data were identical. The response rate for those receiving the gynaecological cancer (GS) questionnaire was 49% and it was 71% for the breast cancer (BS) questionnaire. Detailed comparison between the two samples revealed several EORTC QLQ-C30 items showing a clear difference in distribution of scores between them. Because of this and the possible bias due to the relative low age-related response rate in GS, only the results from the BS are used for constructing norms. The norms cover all 30 single items on the EORTC QLQ-C30 and the nine derived scales, for women in four 10-year age groups commencing at 30 years and for those aged 70-75. Clear trends in, for example, declining ability to undertake strenuous activity are illustrated and quantified. Levels of certain symptoms, such as pain, are surprisingly high although it is recognized that the population sampled will contain a proportion of women with active disease including cancer. We recommend the use of these norms both as an aid to the clinical assessment of an individual patient, and to assist in the interpretation of clinical trial and longitudinal quality of life data. As a secondary result, we note that a population-based sample will have a lower response rate to a questionnaire with more questions, especially if many of these extra questions are on sexual issues.