The introduction of IFRS 3, which regulates business combinations, replaces the old IAS 22 in 2004. The new standard prohibits the amortization of goodwill and introduces an impairment test under IAS 36. The new subsequent valuation approach is a compromise approach according to which the value of goodwill may or may not be impaired. The impairment test should be performed once a year or when there is an indication for impairment. This approach is characterized as a controversial approach due to its complexity and potential subjectivity that can distort the true picture of financial statements and indicators. The subject of this research is the analysis of variables that affect the amount of impairment losses on goodwill on the example of Croatian listed companies. The aim of this research is to determine the existence of the influence of two groups of variants: numerical (company size, share of financing with long and financial results) and qualitative (change of management and COVID). The sample consisted of 40 companies from the Zagreb Stock Exchange, which in the period from 2015 to 2021 showed goodwill in their balance sheets. The results of the analysis show that the companies in the sample very rarely reduce the value of goodwill. Out of a total of 190 observed observations, only 26 observations showed a decrease, shown as a percentage, only 13.68% of the observations showed a decrease in goodwill. The influence of variables was examined by a logistic regression model with a fixed effect. The model determined a statistically significant impact on the amount of impairment loss in financial results and the size of the company. These variables have a negative impact, in other words, an increase in the ratio of the financial result to the value of the active one can be expected to reduce the probability of a decrease in goodwill.