<p>This study assessed soil organic carbon (SOC) content and its spatial variability at 130 agricultural sites of an inter-Andean valley in the Guachal watershed in southwestern Colombia. The research tested two hypotheses: (1) the SOC content varies significantly across orders, types, and environmental conditions and (2) geostatistical methods incorporating auxiliary variables improve the accuracy of SOC prediction. Ordinary kriging (OK) and cokriging (CK) techniques were used to map SOC based on soil properties and environmental variables. The content of SOC varied from 20 to 294 t ha⁻¹. The CK performed better than OK by including additional variables such as total nitrogen, pH, and Fe, reducing prediction error and improving spatial accuracy. The highest SOC content (174.83 t ha⁻¹) was observed in soils from the order andisols located on steep slopes (50–75%) and in cold climates (12–18&#xa0;°C) with secondary vegetation. Soils from the order mollisol, located in flat areas, with temperatures above 24&#xa0;°C and covered by cultivated pastures, had the lowest organic carbon content (20.42 t ha⁻¹). The major agricultural crops (sugarcane and cocoa) showed minimal variability in SOC reserves. These results indicate the critical role of soil order, land use, and environmental factors in SOC distribution. Furthermore, they provide essential information for policies aimed at improving carbon sequestration and sustainable soil management in tropical agricultural landscapes.</p>