OBJECTIVES: The Centre for Epidemiologic Studies Depression Scale (CES-D) is one of the most widely used depression questionnaires; it has been translated into many languages and is frequently used in multi-national studies. This research sought to examine whether different language versions of the CES-D were affected by country (cultural)-related differential item functioning (DIF). METHODS: CES-D data were available from depressed patients in the UK (n = 177), US (n = 100), Germany (n = 78) and France (n = 124). The data were pooled and applied to the one-parameter Rasch item-response model for analysis and to identify cross-cultural DIF. RESULTS: The UK and German CES-D did not fit the Rasch model (Chi2 p<0.001) suggesting that summation of item scores in these countries is not justified. Four items in the UK (including 2 of the 4 positively worded items) and 2 items in Germany misfitted. The US CES-D exhibited borderline overall misfit to the Rasch model (Chi2 p<0.01) with no item misfit and the French data fitted the Rasch model (with 1 item misfitting). The pooled data from the 4 countries did not fit the Rasch model (Chi2 p<0.001) and DIF was observed in 7 items (including all of the positively worded items). DIF between the US and UK (5 items), the US and Germany (5 items) and US and France (4 items) was greater than that between UK and Germany (1 item), the UK and France (2 items) and Germany and France (3 items). CONCLUSIONS: CES-D data from these countries cannot be pooled justifiably without first accounting for DIF by culture. DIF appeared to be greater between the US and Europe than within European countries. In addition, the use of both positively and negatively worded items in a questionnaire may introduce bias.