This study aims at identifying determinants of health related quality of life in Poland, and in particular at verifying whether health domains are complements or substitutes and what the impact of heterogeneity of population on the health state valuation is. The paper uses data in panel structure coming from a survey conducted in Poland and consisting of 6700 valuations (after data cleaning) of EQ-5D health states with time trade-off method. Several econometric models are built in order to detect the impact of complementarity and heterogeneity. Random effects models as well as random parameters models estimated using Bayesian approach are used. The results show that health domains are complementary goods. Especially the lack of pain/discomfort is a complement to other health domains. Demographic factors influence how health state change impacts utility. These factors encompass sex, education, respondent’s health state and even belief in life after death.