The authors attempted to determine the effects of a high-stakes standardized testing-driven instructional approach on mathematical performance. The authors developed a multiple-choice mathematics performance test for 1,006 Grade 4 students in 28 North Cyprus schools. Analysis revealed that students who spent more time on test-taking skills performed better, especially in routine mathematics items, than did students who spent less time on test-taking skills. There was no difference observed in test results from nonroutine story problems. However, analysis did indicate that spending too much time on test-taking skills led to memorizing procedures and cuing on surface attributes of a problem.Key words: high-stakes standardized tests, routine and nonroutine mathematics items, test-taking skills