Rapid and accurate acquisition of crop coefficient (Kc) values is essential for estimating field crop evapotranspiration (ET). The lack of rapid access to the high-resolution spatial and temporal distribution of Kc values hinders obtaining a crop Kc value for application in precision irrigation agriculture. This study aimed to explore the potential of leaf area index (LAI) and multispectral vegetation indices (VIs) obtained by an unmanned aerial vehicle (UAV) for estimating the Kc value for a maize crop on a field scale and to obtain a high-resolution spatialtemporal map of Kc values. Hence, the performance of the estimation model for daily maize Kc derived by two machine learning algorithms (random forest regression-RFR and multiple linear regression-MLR) based on the ground-based LAI and six types of UAV-based multispectral VIs (normalized difference vegetation index, NDVI; soil adjusted vegetation index, SAVI; enhanced vegetation index, EVI; transformed chlorophyll absorption in reflectance index, TCARI; green normalized vegetation index, GNDVI; and visual atmospheric resistance index, VARI), was evaluated under multiple irrigation conditions during the entire cropping cycle. Maize RFR with VIsLAI-based ET was compared to soil water balance (SWB) and FAO-56-based ET. The results showed that the RFR algorithm effectively (R2 = 0.65) estimated maize Kc values based on ground-based LAI and UAV-based VIs. The UAV-based VIs based on Red-edge-Red and Green-Red spectral bands and ground-based LAI were suitable predictors in the Kc prediction model under different irrigation conditions. Further, we successfully obtained a high resolution (pixel size of centimeter) spatial distribution of maize Kc values based on EVI-based LAI and UAVbased VIs. Furthermore, the results indicated that the combination of UAV multispectral remote sensing technology and the RFR algorithm provides a potential solution for the distribution of water use and precision irrigation on a field scale.