OBJECTIVES: To cluster health status among adults with hemophilia by applying the hidden Markov model (HMM) to account for longitudinal changes in quality of life scores and to derive utility weights for each cluster. METHODS: Data were obtained from the Hemophilia Utilization Group Studies (HUGS), a prospective, multi-center observational study conducted from 2005 to 2013.  Demographic and clinical characteristics, bleeding frequency and health-related quality of life (SF-12) at initial interview and 3-month follow-up visits for 2 years were collected. This analysis included data for 211 adults with hemophilia A or B with at least two observations. The HMM was adapted to generate the clusters and to fit the observation sequences, which consisted of SF-12 physical component scores (PCS) and mental component scores (MCS). Utility weights were derived using Brazier’s algorithm by mapping SF-12 scores to a SF-6D scale.  RESULTS: Mean age was 34.6 ±13.2 years and 59% had severe hemophilia. Data in four unique clusters provided the best fit with the model. Mean PCS were [54.2±2.5, 57.1±2.4], [50.4±6.5, 47.3±9.6], [32.4±7.7, 54.4±7.1] and MCS were [35.3±7.1, 36.8±8.6] from cluster 1 to cluster 4, respectively. Utility weights for each cluster were 0.91±0.05, 0.75±0.11, 0.65±0.11 and 0.55±0.08, respectively. At baseline, 49 (23%) adults were classified in cluster 1, 63 (30%) in cluster 2, 67 (32%) in cluster 3 and 32 (15%) in cluster 4. Being in a worse health cluster was significantly associated with unemployment, low household income, having severe hemophilia and higher BMI. The mean bleeding frequencies during a 3-month period were 1.54±2.62, 3.01±4.91, 3.80±4.09 and 4.83±5.69 among those who were assigned to cluster 1 to 4, respectively (p<0.0001). CONCLUSIONS: Health status in hemophilia can be described by four mutually exclusive and clinically relevant clusters, which provide patients’ value of health outcomes for future economic analyses.