OBJECTIVES: The SF-36 is a measure of health related quality of life that is widely used internationally.  SF-36 produces scores for eight dimensions (physical functioning (PF); role physical (RP); bodily pain (BP); general health (GH); vitality (VT); social functioning (SF); role emotional (RE); mental health (MH).  However there is debate about whether these dimensions are applicable cross culturally, and also the relationship between the MH and VT dimensions. The aim was to assess the dimensionality using multinational datasets as part of the development of SF-6D-V2. METHODS: Exploratory and confirmatory factor analysis was used to examine SF-36 dimensionality, and was applied to patient and general population datasets from the UK, Australia, Canada, USA and Japan (n = 55,923). Analysis was carried out separately for each country, and on the combined data. The general health items were not included as the focus was the specific health areas measured by SF-36. RESULTS:  Exploratory factor analysis on the data from the English speaking countries suggests that the PF, RF, BP, SF and RE dimensions are generally consistent but there are inconsistencies regarding the MH and VT, where the items split into factors based on whether the item is positively or negatively worded. The data from Japan suggests that the role dimensions do not split into physical and emotional constructs. Confirmatory analysis suggests that both the original seven factor model, and a model splitting MH and VT based on the direction of the items, fit the data acceptably. CONCLUSIONS: There is evidence for cross cultural differences in the role functioning dimensions of the SF-36, most likely due to differences in the perception of emotional health. The inconsistency of the MH and VT dimensions may be due to the combination of positive and negative items, order effects, or content overlap.