Testing is a critical tool in classrooms to which school authorities, teachers, students and parents benefit, which means that its importance can never be belittled nor despised. This study aimed at identifying the common test types used by university instructors teaching mathematics and the taxonomic level of items they used in the construction of tests. Reasons and the processes of constructing tests were also identified. Qualitative and quantitative data analyses were appropriately used to achieve the research objectives. Findings showed that university mathematics instructors always constructed tests to evaluate students which became bases for grading them, motivating them to study, identifying weaknesses both in teaching and learning, and obtaining feedback about teaching. Instructors believed that test results were perfect proofs of the quality of teaching, progress in students’ knowledge, and sufficiency of motivation in the class. They often used problem-solving and computational tests, which were in applying and analyzing taxonomic levels. Constructing tests was done by going back to topics included in the course syllabi and ensuring that all objectives were addressed in the tests.