OBJECTIVES: To elicit preferences for motor-symptoms, side-effects and process characteristics of competing treatments in Parkinson’s Disease and comparing within respondents a best-worst scaling profile (case 2) and multi-profile (case 3) task. METHODS: Seven attributes were derived from patients' and neurologists' interviews: treatment modality, tremors, slowness of movement, posture and balance problems, fatigue, dizziness, and dyskinesia. A profile (BWS2) and multi-profile (BWS3) best-worst scaling experiment were designed to observe if different types of BWS result in differences in valuation. In the BWS2, respondents were asked to select the aspects that they perceived as most and least preferable within one single treatment profile. For BWS3, respondents were asked to select the most and least desirable treatment from three profiles. Nine BWS2 and ten BWS3 tasks based on a D-efficient design were offered to 596 respondents from an internet panel. The data was analyzed with mixed conditional logit. For six predefined states, including best and worst, estimated utility values were calculated. RESULTS: The attributes' rankings based on the relative importances differ slightly between tasks. Dizziness is most important attribute in BWS3 and tremors in BWS2. There was high agreement between estimated coefficients (r=0.97). The slope between the coefficients is less than unity, suggesting that overall the BWS3 coefficients are scaled up to a larger extent than the  BWS2 coefficients. This implies smaller residual variance and lower decision uncertainty with the BWS3 task. The estimated utilities' rank order for the six states is the same for 64.4% of the respondents. One rank reversal occurs for 29% of the respondents. CONCLUSIONS: There is agreement on attribute importance and utility estimates both within and between respondents. There is yet no conclusive evidence for one of the two tasks, they both have their merits and depend on the context. The choice between them should be based on pretesting results, study goal and target population.