The geological storage of CO2 is a very promising technique to minimize the effects of climate change. In Brazil, the pioneering project of geological storage of CO2 will complement activities of enhanced oil recovery in mature fields of Recôncavo Baiano. This dissertaion aims to analyze the data of an experimental program that sought to establish the distribution of carbon dioxide (CO2), methane (CH4), radon (Rn), microbiota and light hydrocarbons (C1-C5) in an area of 30km2 in an oil field in the recôncavo baiano field as part of the establishment of environmental control of the geological storage of CO2 program. Analysis of the data contemplated a descriptive classic statistical study and a geostatistic study to assist in interpretation of the experimental program data and mapping the distribution of these parameters for the oil field. The descriptive statistics analysis presented classical distribution, dispersion and position parameters. Through these values was possible to conclude that the distributions of CO2, CH4, NMP, Rn and some values of study presented heterogeneity and asymmetry. The homogeneous distribution was observed for values of temperature, moisture content, porosity and some light hydrocarbons (i.e., propene, 2-Butene Trans, 1 Butene, 2-butane-Cis and N-Pentane). The data did not fit the normal distribution, except by the values of soil moisture content. Geostatistics codes were used to generate data distribution maps, in order to better visualize the local distribution of concentrations of the parameters of interest. It was found that, among the classic interpolation methods, that best fit to the data were obtained by the method of inverse distance and krigging. By analyzing semivariograms was possible to observe the occurrence of spatial variability of the data. It was noted a spatial dependence for the values of CO2 and CH4. However, the spatial dependence of Rn and Microbiota proved less expressive. Spatial variability of soil properties and parameters observed in this study reveals that soils are highly structured spatially and that such a condition must be taken into consideration when choosing the most appropriate experimental methodologies for monitoring programs and future sampling.