A revision of the Quality of Life Questionnaire (QLQ-C30) of the European Organization for Research and Treatment of Cancer (EORTC) was undertaken to improve low internal consistency estimates (Cronbach's alpha) and content validity for the role functioning scale and a conceptual difficulty (undue emphasis on physical functioning) in the global quality of life (QOL) scale. The role functioning items were reworded and a four-category response format was substituted for the previous dichotomous format. A new item asking about 'overall health' was substituted for the 'overall physical condition' item in the global QOL domain. The original and new versions were tested at three time points in a total of 1,181 patients with cancer in Canada (n = 696) and the Netherlands (n = 485). In both samples there was a marked improvement in internal consistency for the role functioning scale (Cronbach's alpha s ranging from 0.78-0.88) in the new version. In the global QOL scale, the substitution of the new item for the previous one did not alter internal consistency (Cronbach's alpha's ranging from 0.81-0.92). The revised versions of the role functioning and global QOL domains have been incorporated into the QLQ-C30 (version 2.0).