Abstract: Soil moisture is a key variable affecting agricultural production and ecological environment. It is therefore of great significance to acquire accurate spatial distribution information of soil moisture. Compare to conventional point-scale based measurements, the Cosmic Ray Neutron Probe (CRNP) provides an alternative approach for capturing soil moisture dynamics to address drought monitoring, plant water stress detection, as well as various hydrological applications. CRNP can realize continuous observation of soil moisture at the field scale and estimate average soil water content with a 0.1–0.2 km² quasi-circular footprint by monitoring the neutron intensity near the ground. However, previous studies conducted soil moisture observation using CRNP are often limited to a single landscape, its ability to observe soil moisture in different geographical environments and vegetation cover types still needs to be further understood. In this study, data from 16 representative stations located in different climatic zones and under different vegetation cover conditions worldwide were selected, including seven grasslands covered stations, three farmlands covered stations, three forests covered stations and three shrubs covered stations, and a unified data processing procedure and soil moisture estimation method were used to invert the average soil moisture observation value in the observation area. The environmental factors affecting the neutron data were corrected, including atmospheric pressure correction, air humidity correction, incident neutron intensity correction and aboveground biomass correction to remove their influence to neutron counts. The N₀ parameter method was used to invert the soil water value. Ground point-scale based soil moisture measurements within the CRNP footprint were aggregated to validate the soil moisture estimates based on CRNP observations. Therefore, a systematic and comprehensive comparison of soil moisture observation capabilities of CRNP over different climatic zones and vegetation cover conditions can be performed. The results indicate that, in general, CRNP observations under different vegetation covers and vegetation cover conditions were highly sensitive to soil water dynamics and can clearly respond to precipitation events. However, there were significant differences in the estimation accuracies. CRNP outperforms over grasslands, except result of one station was relatively poor, the other stations showed high reliability, , with the optimal RMSE of 0.05 cm³/cm³, the optimal ubRMSE of 0.03 cm³/cm³, the optimal Bias of 0.006 cm³/cm³, and the optimal R² of 0.88. In shrub covered areas, the CRNP observation effect of selected sites was strongly affected by the local environment, and the accuracy of soil water estimation is quite different. In was shown R² value varies greatly, from 0.91 at the best to 0.02 at the worst, while ubRMSE, RMSE and Bias are repectively of 0.01 cm³/cm³, 0.02 cm³/cm³, and -0.02 cm³/cm³ at the best condition. In terms of the CRNP sites under grassland cover, the differences in CRNP observations are mainly related to the growth of surface vegetation, etc.; at sites under crop cover, the differences in CRNP observations are mainly related to anthropogenic activities in the agricultural areas and surface heterogeneity; at sites under forest cover, the differences are mainly related to the high canopy, etc. With repsect to the sites under shrub cover, the differences are mainly related to climate, etc. The Yanco and Sidaoqiao sites have the worst observational accuracy among all the sites, with an R² of only 0.46 at Yanco and 0.02 at Sidaoqiao, which have weak data consistency may be mainly related to special local soil moisture environments. In conclusion, CRNP observations of soil moisture are more likely to capture regional moisture fluctuations than point-scale in situ observations. This study helps us better understand the capabilities and potentials of CRNP in observing soil moisture, and provide a scientific basis and reference for applying CRNP observation to capture the spatio-temporal changes of soil water at field scale under different vegetation cover and climate conditions.