Noise pollution is one of the most important factors in the deterioration of the quality of life in cities. Natural vegetation in urban green spaces (UGS) is an effective instrument for reducing road noise. The primary pollution field and the parameters of the UGS determine their different ability to reduce noise levels to comfortable or at least safe levels. One of the directions in the field of recreational research is to assess the effectiveness with which UGS provide noise reduction. Accordingly, the purpose of this paper is to provide such an assessment on the example of the UGS of Kyiv. In the study, the primary field of noise pollution from roads with different traffic intensities was determined. Modelling has shown that at the border of green spaces, noise values of up to 73 dB are observed in more than half of the city's UGSs. When determining the effectiveness of the noise reduction function, it was found that the majority of UGSs (75%) have a strong potential to reduce noise levels (from 0.7 to 0.8). The modelling results showed the high efficiency of green spaces in Kyiv in performing this function - 90% of the UGSs provide safe conditions for staying. Converting the efficiency values into volumes of ecosystem services (ES) based on the Harrington's desirability function allows us to understand that, given the current levels of traffic load and the parameters of UGS, the maximum volumes of noise reduction ecosystem services provided by vast majority green spaces in Kyiv (2982 objects).