The aim of the study was to validate a Finnish version of the internationally used Western Ontario and McMasters University osteoarthritis index (WOMAC) questionnaire, by testing its reliability, validity and responsiveness. The subjects of the study were patients scheduled for elective total knee arthroplasty (TKA) or total hip arthroplasty (THA). The patients completed the WOMAC questionnaire twice prior to surgery and once postoperatively to enable reliability, construct validity and responsiveness analysis. Test-retest reliability could be calculated for 67 patients (37 TKA and 30 THA). The intraclass correlation coefficients (ICC a) exceeded 0.9 with the exception of the stiffness subscale which averaged 0.8. Correlation of the preoperative WOMAC total score with the Harris Hip Score (HHS) in 30 patients and American Knee Society Score (AKSS) in 37 patients were mediocre ranging from 0.48 to 0.53. To test responsiveness, the first preoperative and the postoperative WOMAC scores of 74 subjects were compared (41 TKA and 33 THA). Standard response means (SRM) and effect sizes (ES) exceeded the recommended score of 0.8, ranging from 1.15 to 2.35. In conclusion the Finnish translation of the WOMAC questionnaire performs as the original, is valid and can be used in future studies of osteoarthritis.