The estimation of petroleum reserves entails complex assumptions and calculations, once there is uncertainty associated with volumes, recovery, evelopment and marketability of resources. These uncertainties can be taken into account when a probabilistic approach is employed for the reserves evaluation of each project. Projects are than aggregated to obtain company´s consolidated reserves. The aggregation when conducted probabilistically generates what is known as portfolio effect, which represents the reduction in the standard deviation of the distribution of aggregate reserves. In the context of reserves evaluation, the portfolio effect can increase consolidated proved reserves. However, different exploitation projects may share similar or the same risks, hence the correlation among them is a key factor for estimating more accurately the aggregated reserves. This study proposes a correlation-based methodology for estimating oil reserves probabilistically and is in accordance with the guidelines of the SPE-PRMS (2007) and the Securities Exchange Commission (SEC-2009). The methodology was presented in two major steps: economic assessment considering uncertainty in cash-flow components and probabilistic aggregation using the AHP method partially to determine the correlations among projects. To implement the methodology and run a real data case study a system was created. The numerical results proved the necessity of identifying the correlations among projects and the impacts of the portfolio effect in probabilistic aggregation.