Many examinations with thousands of participating students are organized worldwide every year. Usually, this large number of students sit the exams simultaneously and answer almost the same set of questions. This method of learning assessment requires tremendous effort and resources to prepare the venues, print question books and organize the whole process. Additional restrictions and obstacles may appear in conditions similar to those during the COVID-19 pandemic. One way to obviate the necessity of having all the students take an exam during the same period of time is to use a computer-assisted assessment with random item selection, so that every student receives an individual set of questions. The objective of this study is to investigate students' perceptions of using random item selection from item banks in order to apply this method in large-scale assessments. An analysis of the responses of more than 1000 surveyed students revealed that most of them agree or completely agree with using the proposed method of assessment. The students from natural science departments showed more tolerance of this method of assessment compared with students from other groups. Based on the findings of this study, the authors concluded that higher-education institutions could benefit from implementing the above mentioned assessment method.