Burnout poses severe health‐related and financial risks. However, valid and reliable measurement of this occupational phenomenon has been impeded by the conceptual, psychometric and pragmatical shortcomings of the extant burnout instruments. The Burnout Assessment Tool (BAT) is a new measure of burnout that was developed to overcome these deficiencies. The purpose of this study was to validate the Croatian version of its short form, BAT‐12, using a representative sample of the Croatian working population. To do so, we examined the factor structure and measurement invariance of BAT‐12 across gender, age, and occupational type, as well as convergent and discriminant data on BAT‐12 vis‐à‐vis an alternative burnout measure, the Oldenburg Burnout Inventory, job resources, job demands and job outcomes. The results obtained on online survey data from 966 employees supported the hierarchical structure of BAT‐12, although with slight modifications. We also found comparable loading structure across age, gender, and occupational type, whereas the intercept invariance depended on the moderator. Convergent and discriminant validity was supported in relation to all examined variables. Therefore, the psychometric soundness of the Croatian version of BAT‐12 adds new evidence to the current validation process of the BAT‐12 and supports the application of this measure on the Croatian working population.