Databases usually consist of information drawn from drill hole and similar data. These data very often present some problems which can be solved by using mathematical and geostatistical methods. The present paper deals with the problems of estimating the proportions of lithological types within selective mining units, the problem ofan addition of new drill-holes in an explored deposit and the results in the estimation of its quality or quantity and the problem of safe sampling of gold deposits (sentence is too long). For the first issue, the paper gives an overview of how indicator kriging can be used to make estimates of the proportions of the various lithology within blocks of ground, given the lithology's occurring in the samples. The paper deals with and answers the questions that arise when evaluating an ore deposit after the completion of a drilling or sampling campaign. With the aid of geostatistics one can justify or reject a supplementary sampling program at the same deposit, with all the consequences of such a decision. One of the problems of the sampling procedures during a mineral exploration campaign is the size and the weight of the samples to be taken. The sample proportions matter especially in case ofa gold deposit, due to the rarity and the high value of the mineral. Using the modifiedGy's sampling formula, this paper presents the mathematical considerations and transformations to eventually produce a mathematical formula that enables the exploration personnel to safely sample a deposit of valuable ore.