Balch (1989) reported that students scored higher on multiple-choice exams when item order was sequenced (S) than when it was random (R). To provide cross-validation and to determine whether test anxiety interacts with order formats, three experiments were conducted with a total of 253 introductory psychology students. In two experiments, students were randomly assigned S-or R-format tests. A third experiment used a within-subjects design and a two-part testing procedure with counterbalanced S and R formats. A significant Anxiety × Item Order interaction was detected in one experiment, but no evidence was found for superior performance on S-format tests in any of the three experiments. Concern with practical ramifications of Balch's findings may be unwarranted.