Large-scale land use conversions to support an ever-growing population have occurred around the world in recent years. As a result, parameters such as vegetation type, air temperature, relative humidity, soil moisture and wind movement were significantly affected. Many studies have examined the impact of urbanization on temperature or relative humidity, but few have explored whether or not potential evapotranspiration (PET) is affected by changes in population density and urbanization. In this context, in the present study we attempted to find answers to three key questions. Is there a relationship between PET and urbanization? If so, which parameter, capable of modifying the PET trend, is the most sensitive to urbanization in a given region? Finally, is it possible to measure the impact of urbanization on PET without applying any of the subjective models currently available and used worldwide to solve similar problems? By analyzing recent literature, we found that there is indeed a distinct relationship between PET and urbanization. The most important parameter affected by urbanization which also affects PET was studied by applying multi-criteria decision making and this was found to be air temperature. Finally, a neurogenetic model was applied to reveal the vulnerability of evapotranspiration to urbanization in four cities with different levels of population density. The results confirmed that vulnerability is greater in cities with higher population density and higher levels of urbanization.1. IntroductionMore than 50% of the world's population lives in cities, and the urban population is growing at a much faster rate than the earth's population as a whole and with larger annual increases t...... middle of paper ..... . factors affecting PET are also affected by changes in urbanization. Taking this finding a step further, we were interested in establishing the most important parameter influenced by urbanization that also causes the greatest rate of change in PET. To achieve this objective, a multi-criteria decision-making approach (MCDM) and an artificial neural network (ANN) were used. After introducing MCDM and ANN in more detail in section 2, as well as the specific methodology of the study, the results and discussion are presented in section 3. Finally, the conclusion of the study is reported in section 4.