I describe the motivation, structure, implementation, and student feedback of my mastery-based testing (MBT) system with a significant final exam component in Calculus I and II since 2016. Grades for each written assignment, quiz, and exam question were assigned on a Mastery/Progressing/Needs Improvement (M/P/NI) scale. Homework and quizzes could be reattempted outside of class, and exam topics could be reattempted on future exams, including the final exam. Course grades were determined by the number of M and P grades students achieved in each of the five components WebAssign, Written Assignments, Quizzes, All Exams, and Final Exam by the end of the semester. My MBT system differs from most in that student performance on the final exam counts toward "both" the All Exams "and" Final Exam components. Nearly all student performance on the final exam improved or maintained, rather than lowered, their course grades, suggesting cumulative comprehension with the material. Student feedback was consistently more positive than negative in both courses.