AIM: The main objective of the research is to translate the original English version of the PBPI Questionnaire (Pain Beliefs and Perceptions Inventory) into Croatian and to test its psychometric properties in patients with chronic pain in Croatia and to evaluate its reliability, validity and usability – validation of the questionnaire. METHOD: 131 participants were surveyed in this study. The participants were patients with chronic pain of different etiologies and with different types of treatment cycles being treated in the Clinic for Pain Management of KBC Split. Of the total number of surveys collected, 101 were valid. The measuring instruments used in this study are the sociodemographic questionnaire, the general pain questionnaire at the beginning of the therapy cycle, the general pain questionnaire at the end of the therapy cycle, and the Croatian version of the PBPI questionnaire (Pain and Perceptions Inventory; Williams and Thorn, 1989.). The PBPI questionnaire consists of 16 statements. RESULTS: The VARIMAX factor analysis procedure for the Croatian version of the PBPI questionnaire defined 3 factor solutions of questionnaire „Pain Beliefs and Perceptions Inventory“ („TIME“, „MYSTERY“ and „SELF – BLAME“). The Cronbach alpha indicators were determined for the „SELF – BLAME“ component (at the beginning of the cycle is 0.788; at the end of the cycle is 0.882), „TIME“ (at the beginning of the cycle is 0.704; at the end of the cycle is 0.784), and „MYSTERY“ (at at the beginning of the cycle is 0.649; at the end of the cycle is 0.725) indicating that the measuring instrument is valid. CONCLUSION: The Croatian version of the PBPI questionnaire is applicable and reliable for measuring pain perception and belief in patients with chronic pain in Croatia. „TIME“ was positively associated with the assessment of pain intensity at the beginning (ρ=0.24; P=0.016) and at the end of the therapy cycle (ρ=0.44; P<0.001). It was found that at the end of the therapy cycle, there was a strengthening of association as the correlation coefficient (ρ) recorded an increase in value. The other two components of the questionnaire were not positively associated with the assessment of pain intensity. Comparing the results at the beginning of the therapy cycle and at the end of the therapy cycle, we found that there was a decrease in the level of pain (Z=7,90; P<0,001).