Within the framework of studying the public sector and institutional entities cities have one of the most important roles in the functioning of the public sector, as well as in the maintenance of the public system. The importance of cities in the Republic of Croatia stems from the fact that in year 2018 cities (127 cities and the City of Zagreb as a unit with a special status) spent HRK 17.9 billion on their expenditures, what, in relation to state budget expenditures (129.6 billion in the same year), indicates the financial capacity and the role of cities in the public sector. Regardless of secure and stable sources of financing during 2018, 116 cities in the Republic of Croatia had HRK 684 million in overdue, unpaid liabilities. Determining the factors that affect the inability to pay due obligations, e.g. default factors, of cities in the Republic of Croatia is the main objective of the analysis and research of this doctoral thesis. Because of the fact that in the Republic of Croatia there is no legal possibility of bankruptcy of cities and that bankruptcy proceedings are most often carried out due to long-term inability to pay obligations, further research on variables that affect the bankruptcy of cities was not possible in this doctoral thesis. The conducted research used a replacement (dependent) variable created through the ratio of overdue, unpaid liabilities and total expenses and expenditures (default_proxy). Based on previous scientific knowledge and available data for all cities in the Republic of Croatia, a set of data was created for 33 independent variables, which are divided into three groups (finance and budget, liquidity and debt, population), and whose impact was tested on the dependent variable in time period from year 2014 to year 2018. A key feature of this study is in monitoring the dynamics of the movement of different variables over a five-year period. Using different statistical methods, seven statistically significant factors were singled out, creating a multiple regression model based on three independent variables. The variables of transferred revenue shortfall, of operating income receivables and the ratio of current assets to current liabilities have the greatest strength in explaining the variations of the dependent variables. The predictive ability of the model is significant, and the selected independent variables of the multiple regression model explain 60.1% of the variations of the dependent variable. The presented equation for calculating the future value of the average ratio of overdue liabilities in total expenditures and expenditures (default_proxy ') provides a simple application solution that can serve as an early warning system for all cities, but also for the relevant ministryand creditors. In order to reduce overdue, unpaid liabilities of cities, it is necessary that city leaders manage their liquidity and budget funds more rationally, that the legislature and the executive improve the existing legislative framework, and finally, that credit ratings of cities are created and used. Also, more up-to-date and comprehensive, primarily economic, data will enable future researchers to refine the presented model further.