We present a cost-effective statistical approach that integrates satellite imagery, environmental variables and ground inventory data to map the spatial distribution of aboveground woody biomass suitable for charcoal making. The study was conducted in the Cuitzeo basin located in central Mexico, where charcoal is produced from oak forests covering approximately 10% of the total area (4033 km2). Diameters of trees and sprouts in 78 plots of 0.2 ha each was measured. Allometric equations previously developed locally that only require tree diameters were employed to estimate the amount of woody biomass suitable for charcoal making i.e. the amount of wood that is loaded into the kilns. The performance of two statistical techniques for the interpolation of field data was assessed by cross-validation; these techniques were linear regression and regression-kriging, the second taking into account the spatial autocorrelation of data. Spectral bands, vegetation indices, texture measurements and variables derived from a Digital Elevation Model were examined as explanatory variables. Accounting for spatial autocorrelation (regression-kriging) improved the model's R2 from 0.61 to 0.69, representing a relative error reduction of 11.3% (from 11.01 to 9.77 t ha− 1 of wood suitable for charcoal). The available stock was compared to current estimates of charcoal demand in the Cuitzeo basin and insights were given on how this information can be used to estimate the annual sustainable production potential of oak in order to account for supply–demand balances.