The possibility of analyzing multidimensional data for predicting properties of petroleum road asphalts is examined. Construction of a prognostic model helped determine the optimal feedstock composition and technological parameters of the oxidation process for getting improved asphalts. The positive and negative correlating variables, viz., feedstock properties, oxidation process parameters, and properties of oxidized asphalts, are determined from the data set. The analysis results were used to develop a technology for producing improved asphalts under conditions of asphalt production by OAO Syzranskii NPZ (OJSC Syzran Refinery).