The Zambujal ore deposit, Neves-Corvo mine, is a zoned volcanogenic sulfide deposit of copper and zinc, with massive ores at the top and stockwork ores at the bottom. Metal grades are strongly zoned by ore types. The main methodology hereby proposed combines geostatistical techniques so that an improved stochastic geological model of the Zambujal deposit encompassing morphology and grades is presented. The model of the morphology was made in two main steps. First, a 3D solid of the boundaries was created and then a 3D grid model of the local sulfide proportion was simulated. This latter variable was modeled by using rock specific gravity as a proxy. After that, a conditional grid model of relative copper grades, i.e., recalculated metal grades assuming only the sulfide content, is also simulated in accordance with the morphology. At the end, the new tool, metal tonnage cut-off surface, is proposed, which combines copper grades within massive ores and stockwork ores. To validate the results found, the global tonnages of copper obtained by Ordinary Kriging and the proposed methodology are compared.