This article describes three experiments comparing paper-and-pencil tests (PPTs) to computer-based tests (CBTs) in terms of test method preferences and student performance. In Experiment 1, students took tests using three methods: PPT in class, CBT in class, and CBT at the time and place of their choosing. Results indicate that test method did not affect performance and students liked the idea of additional CBTs. In Experiment 2, students chose the test time and administration method for four exams. CBT was the preferred method and CBT performance did not differ from a PPT comparison group. In Experiment 3, students chose the test method, but not the time, for three exams. PPT was the preferred method. Possible explanations for these findings are discussed. (Contains 1 table.)