OBJECTIVES: The objective of the study is to assess the effect of each stage in the accepted translation methodology for PRO measures (Wild et al. 2005): 2 forward translations and their reconciliation, 2 back translations, back translation review, developer review, harmonisation meeting, linguistic validation interviews and 2 proof readings. METHODS: The questionnaire chosen for this study was the Anti-Clot Treatment Scale (ACTS) which had been translated into Canadian French, French and German. Leaving aside the forward translation and back translation stages, where the whole text is changed, the grid for each stage was reviewed, and changes counted. The types of changes were also considered, to evaluate what each step contributes to the overall translation process. RESULTS: 1) The reconciliation stage and second forward translation allowed the investigator to reconsider their own forward translation and make improvements; 2) the back translation review caught some items where the translation required improvement; 3) only minor changes were necessary at developer review stage; 4) harmonisation allowed tweaks to be made so that all language versions were conveying the same meanings; 5) very few changes were needed following pilot testing; and 6) proofreading caught grammatical and formatting errors. CONCLUSION: As the translations went through the methodology, improvements to the quality and accuracy were made. The forward and back translation processes iron out problematic wordings. Although in terms of numbers of changes the developer review and pilot testing appear to have little effect, these stages serve as an important check of the previous steps. Without harmonisation, the versions may not have been so close in meaning. Proofreading is an important step in removing formatting and grammatical errors that may have crept in when creating the formatted documents from the grids used for the translation process.