OBJECTIVES: There is no evidence about the appropriateness of the EQ-5D (a generic preference-based measure of health) in Myelofibrosis (MF), a rare but serious bone-marrow cancer. This study aimed to provide psychometric evidence of its appropriateness. METHODS: Convergent validity and responsiveness of the EQ-5D was assessed by comparing it to a condition-specific measure, the MF Symptom Assessment Form (MFSAF) in 48 patients with MF with repeated measurement over 48 weeks using data from the ROBUST study. Convergent validity was based on correlation analysis between EQ-5D utility scores and dimensions (mobility, usual activities, self-care, pain/discomfort and anxiety/depression) and MFSAF total score and symptoms (pain, early satiety, night sweats, itching and bone or muscle pain). Responsiveness was based on change in EQ-5D compared to change in MFSAF using standardised response mean (SRM) and Cohen’s effect size (ES). Moderate to strong correlations (rho>|0.3|) and comparable SRM and ES would indicate that EQ-5D was appropriate.  RESULTS: EQ-5D had poor associations with key symptoms in MF (rho<|0.3|), except for the ‘pain/discomfort’ and ‘anxiety/depression’ health dimensions (rho>|0.4|). SRM and ES at week 4 for EQ-5D was 0.270 and 0.343 compared to SRM and ES of 0.911 and 0.826 for the MF-SAF. A large proportion (15.56%) reported no problems in EQ-5D dimensions at baseline. The MFSAF total score did not show comparable ceiling effect (4.76%). There was however some indication of ceiling effect for some of the individual dimension of the MFSAF indicating that patients in this sample did not all have the common MF symptom. CONCLUSIONS: This exploratory analysis suggests that the EQ-5D’s ability to capture the effect of key symptoms in MF is limited to pain rather than the specific MF symptoms such as night sweats and itchiness. However, results of this analysis need to be interpreted with caution due to the small number of patients.