OBJECTIVES: To compare SF-6D and EQ-5D in both, absolute values and sensitivity to change over time, in osteoporotic hip fracture patients in Poland. METHODS: Data were extracted from prospective study on outcomes of osteoporotic hip fractures in Poland (PolHipQol study). Eligible patients had to be 60 years or more, have low energy femoral neck fracture or pertrochanteric fracture of the femur, absence of severe cognitive dysfunction (measured by Hodkinson's Abbreviated Mental Test Score) and both SF-36 and EQ-5D post fracture measurements avaiable. SF-36 scores were translated into SF-6D utilities using the algorithm developed by Brazier et al. The EQ-5D utilities were based on the European VAS value set. The correlations between preference measures were assessed using Spearman's rank correlation coefficient. Sensitivity to change over one year was evaluated with the standardized response mean (SRM). RESULTS: Post fracture data of 65 patients (mean age 77.8; 54 women) and one year follow-up data of 51 patients were avaiable (9 patients were ceased and 5 lost to follow-up). Mean SF-6D utility decreased from 0.65 (SD 0.13) before fracture (recall method) to 0.49 (0.10) after fracture, and then increased to 0.55 (0.12) at the final follow-up. Mean EQ-5D utility decreased from 0.73 (0.22) before fracture to 0.24 (0.17) after fracture, and then increased to 0.47 (0.23) at the final follow-up. SF-6D and EQ-5D utilities correlated strongly: 0.75 (p<0.001). The responsiveness was moderate for SF-6D (SRM=0.40) and high for EQ-5D (SRM=0.83). CONCLUSION: Both SF-6D and EQ-5D captured worsening and improvements in health over time. However, the use of EQ-5D resulted in larger utility losses and gains and in consequence may result in lower cost-utility ratios. The study showed high responsiveness for EQ-5D and moderate for SF-6D, indicating that EQ-5D is more suitable for use as utility measure in clinical trials in elderly hip fracture patients.