Soil nutrients are crucial for understanding the impact of global changes on terrestrial ecosystems. Carbon (C), nitrogen (N), and phosphorus (P) are required in specific quantities by plants and are related to soil fertility. In the Caatinga, one of the world’s largest and most diverse tropical dry forests, there have been limited studies investigating the factors that contribute to the spatial distribution of organic carbon (OC), N, and P in the soil and, even fewer, those that explored the use of Machine learning (ML) modeling. In this research, we aim to predict the spatial distribution of these properties at depths ranging from 0 to 20 cm in the Caatinga biome and assess the predictive capability of environmental and geographic variables. We used the Random Forest model in Google Earth Engine to forecast maps with a spatial resolution of 30 m. The best result was obtained for predicting P [Lin's concordance correlation coefficient (LCCC) of 0.32 and R2 of 0.25], followed by OC (LCCC of 0.25 and R2 of 0.17), N (LCCC of 0.21 and R2 of 0.12) and C/N ratio (LCCC of 0.14 and R2 of 0.10). The final maps exhibited a consistent spatial pattern, with OC, N, and C/N ratio showing correlation with climatic covariates (air and soil temperature in the first 15 cm deep), topographic data (altitude), and geographic regions (longitude and latitude). The variation in P content was primarily influenced by parent material. We highlight the relevance of ecotones, which recorded the highest average levels of C and N and C/N, demonstrating the importance of these areas for the maintenance and dynamics of these ecosystems.