BACKGROUND: Patient satisfaction surveys are increasingly used by hospitals. Many questionnaires are available, but little evidence exists to guide the choice of the most suitable instrument. OBJECTIVE: To compare the acceptability and patient perceptions of 4 patient satisfaction questionnaires. RESEARCH DESIGN: Randomized trial of 4 satisfaction questionnaires: Picker, Patient Judgment System (PJS), Sequs, and a locally developed Lausanne questionnaire. SUBJECTS: Patients discharged from 2 Swiss teaching hospitals (n = 2850). MEASURES: Response rates, missing data, completion time, and patient ratings of the questionnaire (5-point agree-disagree scale). RESULTS: Response rates were similar across instruments (Picker: 70%, PJS: 71%, Sequs: 68%, Lausanne: 73%; P= 0.27). The Picker questionnaire had the most missing responses (mean per item: Picker: 3.1%, PJS: 1.9%, Sequs: 1.6%, Lausanne: 1.1%; P<0.001) and took the longest to complete (minutes: Picker: 19.3, PJS: 12.5, Sequs: 13.4, Lausanne: 13.1; P<0.001), but the fewest patients indicated that the questionnaire failed to address at least 1 important aspect of the hospital stay (Picker: 28.2%, PJS: 38.8%, Sequs: 39.1%, Lausanne: 28.9%; P<0.001). Patient evaluations of the questionnaires were generally similar; the most favorable assessment was chosen by approximately half of the respondents (average of 10 items: Picker: 46.5%, PJS: 46.2%, Sequs: 47.4%, Lausanne: 48.2%; P= 0.60). Key survey results differed considerably by questionnaire. CONCLUSIONS: No questionnaire emerged as uniformly better than the others in terms of acceptability and patient evaluations. All 4 could be used for patient satisfaction surveys. [authors]