The development of a decision support system to compute performance measures of a project, including the uncertainty associated to the activities duration as well as four different types of precedence relationships is presented. The system generates replicates of the project performance, and in each replication activities' durations are simulated, From them, the expected completion time and the probability that each activity be part of the critical path are estimated, as well as their corresponding measurements of error. A validation of the decision support system was performed by computing the empirical coverage, mean and standard deviation of halfwidths, mean square error and empirical bias for the expected completion time of a given project. It was observed that good coverage and halfwidth are obtained for a small number of replications.