Optimally locating additional drill holes depends on initial data configuration, spatial structure of the variable under study, the number of additional drill holes and shape of deposit. Several approaches have been proposed for this problem using geostatistics and optimization methods, but all of them treat the mineral deposit in 2D. An optimization procedure that is based on genetic algorithm is presented for optimally locating additional drill holes in 3D. A case study in an industrial mineral deposit using Al2O3 % grade illustrates the procedure. The results showed that this procedure is in effect in the case of varying thickness.