PURPOSE: To compare two quality-of-life (QOL) questionnaires for cancer patients, the European Organization for Research and Treatment of Cancer Quality of Life Questionnaire-C30 (EORTC QLQ-C30) and the Functional Assessment of Cancer Therapy-General (Fact-G), on the basis of empirical data. PATIENTS AND METHODS: Two hundred forty-four patients with a diagnosis of breast cancer or Hodgkin's disease completed both the EORTC QLQ-C30 and the FACT-G (German language version) during the same session. Questionnaire data were analyzed on a subscale basis using correlation analysis, canonical correlation, and multiple linear regression. RESULTS: Correlations between corresponding subscales of the FACT-G and the EORTC QLQ-C30 ranged from r = .14 for the social domain (very poor agreement) to r = .66 for the physical domain (good agreement), with r values for the other domains lying between these extremes. Canonical correlation analysis for the two sets of subscales revealed that overall agreement between the two instruments was only moderate (first canonical correlation coefficient r = .85, but overall redundancy less than 40%). Of the five FACT-G subscales, only one, physical well-being, was well represented by the EORTC QLQ-C30 subscales (multiple linear regression, R² = .67). Only three of eight EORTC QLQ-C30 subscales (physical functioning, global QOL, general symptoms) were represented fairly well by FACT-G subscales (R² = .43 to .60). The lowest R² values (< .15; ie, virtually no representation by the other instrument) were found for the FACT-G social well-being and relation with doctors and EORTC QLQ-C30 cognitive functioning subscales. CONCLUSION: For the sample investigated, the EORTC QLQ-C30 and the FACT-G were found to measure markedly different aspects of QOL, despite considerable overlap. Replicability provided, this implies that neither of the two QOL instruments can be replaced by the other and that a direct comparison of results obtained with the two instruments is not possible.