Advanced Placement Calculus examinations were administered to nearly 7,000 students in order to determine the impact of calculator use. Both experimental examinations had two sections. Section I items were designed so that a calculator was not needed, but approximately half of the students were permitted to use calculators. Section II items required a calculator to arrive at the correct solution, and all students were permitted calculator use. Calculator use resulted in a substantial improvement in performance for 3 of the 40 Section I items. The internal consistency reliability of the Section I items was higher for students not permitted calculator use. A pattern of negative delta differential item functioning (D-DIF) values for the Section II items indicated that these items were differentially more difficult for females than the Section I items. The D-DIF value of only one Section I item appeared to be substantially impacted by calculator use. Total test scores were found to be higher for those permitted to use calculators, for males in comparison to females, and for those reporting the use of advanced calculators (capable of graphics or the taking of derivatives) in comparison to those reporting the use of scientific calculators. Three appendixes present questions for Calculus AB and BC tests and responses to the student questionnaire. (Contains 3 tables and 13 figures.) (Author/SLD)