Aim: To translate the Consultation and Relational Empathy (CARE) Measure into Croatian and validate the Croatian version of the questionnaire. Methods: A cross-sectional study was conducted in July 2011 in 8 general practices (GP) in Croatia. Following two stages of translation, back-translation, and pilot testing, the Croatian version of the CARE was tested on 568 consecutive patients. Results: Face validity was high, the number of missing values was low (9, and the internal consistency (Cronbach?s alpha) was 0.77. A principal component analysis of 10 CARE Measure items extracted two components with eigenvalues gt;1. These two components explained 43.6% of the total instrument variance. Conclusion: The Croatian version of the CARE Measure had acceptable reliability and face validity, but its intended component structure was not reproduced and further research is needed to understand its dimensionality.