Purpose – The study examines the bank selection criteria employed by Ghanaian university students. Design/methodology/approach – We used convenience sampling to select 997 students aged between 15-30 years from Kwame Nkrumah University of Science and Technology. Exploratory factor analysis was first conducted to determine the constructs that measure students’ selection of bank criteria. Using binary logistic regression, the extracted constructs were used as independent variable on the bank patronized. The effects of student demographics on the bank selection criteria was also determined using a multiple linear regression. Findings – The study extracted six constructs that measured bank selection criteria by university students. These were operational competence, external influence, physical evidence, e-banking facilities, convenience and cost of operating bank account. Out these, e-banking facility, convenience and cost of operating bank account, were statistically significant at determining the selection of bank. The department students belonged to (social science or pure science) affected the level of weight placed on cost of operation. Age of respondents and department affected the premium placed on e-banking. Finally, employment status and department affected the level of importance student attached to convenience as a selection criterion. Gender of students had no statistical effect on any of the bank selection criteria. Originality/value – The reviewed literature showed that, researchers either explored in isolation, bank characteristics influencing bank selection by clients, or client demographic and preference for bank and its characteristics. This study sought to feel this gap by combining the two, to provide a more robust model in explaining students’ selection of bank.