6519 Background: Patient-reported outcomes (PRO) assess patients' perspectives on the impact of diseases and treatments. There is increasing interest in having patients complete PRO questionnaires and using the data clinically to identify and address patients' issues. There are many different PRO questionnaires, and this study investigated whether one PRO is more appropriate than others for clinical use. Methods: This controlled trial randomly assigned patients to complete 1 of 3 PRO questionnaires: six domains from the Patient Reported Outcomes Measurement Information System (PROMIS), the EORTC QLQ-C30, or the Supportive Care Needs Survey-Short Form (SCNS). Breast or prostate cancer patients, >=21 years old, undergoing treatment, with oncologist visits at least monthly were eligible. Patients completed their assigned PRO questionnaire via computer either in the clinic or via the web prior to their clinic visits for the duration of active treatment. The results were provided to their clinicians for use during visits. After treatment ended, patients completed a 13-item feedback form to assess the impact of the questionnaire on their care (e.g., easy to complete, promoted communication, improved care quality). We hypothesized that there would be no difference in the proportion of patients who strongly agreed/agreed to all 13 items in the feedback form by questionnaire arm. Results: Of 294 eligible patients approached, 224(76%) enrolled (mean age 66, 78% white, 28% breast, 72% prostate). Of these, 181 patients (81%) responded to the feedback form, with 116 patients (52%) completing all 13 items. Of the 116 patients with complete data, subjects in the QLQ-C30 arm were most likely to strongly agree/agree to all items (82%), followed by PROMIS (62%), and SCNS (56%) [p=.05]. Among the 181 patients with at least partial data, the proportions strongly agreeing/agreeing to all items were 74% for QLQ-C30, 61% for PROMIS, and 52% for SCNS [p=.03]. Conclusions: In a randomized trial comparing the benefits of PROs applied in clinical practice, we found significant differences among the 3 PRO questionnaires. Future applications of PROs for patient care should select the most appropriate PRO measures for the purpose and setting.