OBJECTIVES: Intensive cognitive interviews are valuable in the refinement of patient reported outcome (PRO) instruments and can identify how patients perceive the meaning of questions, how they develop and justify their responses and can allow the researcher to test face validity and burden of the instrument. We conducted intensive cognitive interviews with outpatients diagnosed with schizophrenia to refine the clear thinking scale (CTS) -- a recently validated PRO instrument that assesses patient preferences and functioning across four sub-domains (staying organized; making sense of the world; feeling clear headed; and expressing thoughts and feelings). METHODS: Intensive cognitive interviews were conducted by an experienced interviewer at various facilities associated with Johns Hopkins. Following the consenting process, participants completed the preferences (5 items) and functioning (28 items) components of the CTS. Concurrent and retrospective verbal probing was utilized to assess the respondent level of understanding and burden. All interviews were recorded, transcribed and analyzed using phenomenological qualitative methods. RESULTS: A total of 20 patients with a primary diagnosis of diagnosis of schizophrenia or schizoaffective disorder participated in the study. While respondents regularly got off topic and often lacked confidence in their answers, we found a relatively consistent interpretation of items across respondents. We also identified several approaches to improving the CTS scale, including: 1) simplifying items and avoiding compounded items; 2) repeating the time frame (i.e., in the past week) before every question; 3) supporting response variables with numbers/details; 4) using rankings rather than ratings to elicit patient preferences; 5) avoiding items that might have a literal interpretation; 6) measuring functioning prior to measuring preferences; and 7). having an example question for both functioning and preferences questions. CONCLUSIONS: Despite its difficulties, intensive cognitive interviewing among patients diagnosed with schizophrenia provided valuable information for the refinement of the CTS and other future PRO scales.