OBJECTIVES: More than 10 disease modifying drugs (DMDs) are available for relapsing-remitting multiple sclerosis (RRMS) and clinically isolated syndrome (CIS). Poor adherence to DMDs often results in suboptimal (cost-)effectiveness. Understanding patients’ preferences for DMD and incorporating them in clinical decision-making can contribute to adherence to DMDs. This study aims first to identify the full spectrum of DMD attributes and second to quantify their relative importance among patients. METHODS: First, 3 focus groups with RRMS patients were conducted using a nominal group technique to explore attributes for decision-making. Through individual generation and discussion of attributes, a full list was created and patients selected their 10 most important attributes. Second, a best-worst scaling (BWS) was conducted in a larger RRMS/CIS patient group to prioritize attributes. Attributes’ mean relative importance scores (RIS) were obtained with Hierarchical Bayes analysis.  RESULTS: Nineteen patients participated in the focus groups. A list of 34 attributes was derived of which 7 were excluded from the BWS because they were never included in the top 10. Next, 185 patients evaluated the importance of 27 attributes in the BWS. Effect on disease progression was most important (RIS=9.6), followed by quality of life (9.2), relapse rate (7.8), severity of side effects (7.6), relapse severity (7.4), current MS symptoms (7.3), plaque development (7.3), and safety (6.0). Effect on disease progression was 1.8 times more important than influence on lifestyle (5.3). Subgroup analysis showed that DMD naïve patients found side effect-related attributes far more important than DMD experienced patients.  CONCLUSIONS: Using a thorough method, this study reveals that patients value effectiveness and unwanted effects most. Also, heterogeneity was observed in different subgroups. Clinicians should be aware of the average preference values and that DMD attributes are valued differently by different patients. Person-centred clinical decision-making would be needed and requires elicitation of individual preferences.