Ranking of universities on different criteria is a vital problem for modern society. Worldwide Ranking in the educational sphere is based on available open-access sources that often contain imprecise, consumer-oriented information with questionable reliability. It can be noted that information on the ranking of universities usually is uncertain and with partial reliability. Decision-making on this type of information leads to unreliable results. In this paper, we suggest a methodology for rankings on universities, based on three main criteria, mainly academic reputation, employer reputation, and citations per faculty by using the Z-number theory. This approach allows considering the imperfectness of ranking information. The numerical example which includes three different scenarios, shows the effectiveness and validity of the proposed approach.