GradesFreeNo signup
At 87% going into a final worth 20% of the course grade, a 90 needs 102 on the exam. Those are the figures already in the boxes below, and the answer is that it cannot be done: 89.6 is the highest grade that course can still finish at. Put your own numbers in, then read what the answer means when it comes back above 100.
Last updated 1 October 2026
A planning estimate, not a grade. This page solves one equation: your course grade is your average on everything except the final, carrying 100% minus the final’s weight, plus the final itself. It does not know your syllabus. It does not model a dropped lowest score, a final that replaces another test, extra credit, a curve, a participation grade, a late penalty, or your instructor’s rounding policy, and every one of those changes the answer. The grade of record is the one your instructor submits.
Your course grade at the end of term is a weighted average of two things: everything you have already done, and the final. Call your average on everything else C, the final’s share of the grade w, and the final score F. Then:
course = C × (1 − w) + F × w
At 87% going in with a final worth 20%, that is 87 × 0.80 + F × 0.20, which is 69.6 + 0.2F. The 69.6 is already fixed. Four fifths of the course is over and cannot be changed by anything you do in the exam room.
This page runs the equation backwards. Set the course grade to the target T and solve for the final:
F = (T − C × (1 − w)) ÷ w
That is the whole calculator. For a 90 in the example above: (90 − 69.6) ÷ 0.2 = 20.4 ÷ 0.2 = 102.
Two cases, both exact
In the first case the target of 80 sits below the current average of 84, and the required score is still well short of it, because 84% already carries three quarters of the weight: (80 − 84 × 0.75) ÷ 0.25 = 17 ÷ 0.25 = 68. Sixty-eight on the final holds an 80 in the course. The second case is this page’s own starting numbers, and it is the answer most readers actually get.
A required score of 102 is not a warning or a suggestion to study harder. It is a finished statement about arithmetic: no score on that exam, including a perfect one, produces a 90 in that course. The target was gone before you opened this page.
So the useful question is not what you need. It is what is left. Put F = 100 into the same equation and you get the best grade still available:
best possible = C × (1 − w) + 100 × w
At 87% with a 20% final: 69.6 + 20 = 89.6. A perfect final finishes the course at 89.6. If the A− line is 90, it is four tenths of a point out of reach, and no study plan closes it. Put F = 0 in instead and you get the floor, 69.6, which is where the course lands if you do not show up. The entire remaining argument is a twenty-point band, and the final decides where in that band you finish.
The band, not the target
The second pair is the harsher version. At 78% with a final worth 15%, a 90 requires 158 on the exam and the ceiling is 81.3, so nothing in the exam period can produce an A, a B+ or a B. What is being decided is whether the course finishes as a B− or worse, which is a real decision, just not the one the reader came in asking about.
Three things move a ceiling and studying is not among them: extra credit that genuinely exists, a drop or replacement rule you have not accounted for, and a correction to a mark already entered.
The equation needs C to be your average across all of the non-final weight. The large percentage at the top of your course page usually is not that, and the learning-management systems are explicit about why.
Ohio State’s Canvas guidance sets the two figures side by side. “Current Grade is calculated by adding up the graded assignments only according to their weight in the course grading scheme”, while “Total Grade is calculated by adding all the assignments according to their weight in the course grading scheme (both graded and ungraded assignments)”. Of the first it says plainly: “Ungraded assignments are ignored in this calculation”.
The default view is the first one: an average over work that has been marked, with anything assigned but not yet graded absent from it rather than counted as zero. That is the right number to show a student in week nine, and the wrong number for this box, which asks about the whole non-final portion of the course.
Work an example. Homework is 20% of the course across ten assignments. Eight are graded and average 95, two are unmarked, and your current grade reports the category at 95. If those two come back at 60 the category is really 88, so your pre-final average falls by 0.20 × 7 = 1.4 points. At a 20% final each point of pre-final average is worth four points of exam, so the score you need moves by 5.6. You planned against a number that was true and not applicable.
The honest way to fill the box is to enter the average you expect across the entire non-final portion, with your own estimate standing in for the work that has not been marked, and to re-run it when those marks land.
The required score moves by 1 ÷ w for every point you move the target. That single fraction explains most of what feels unfair about finals week.
Exam points needed per extra point of course grade
The sharper companion is how far above your current average the final can carry you at all, which is (100 − C) × w. At 87% with a 20% final that is 13 × 0.20 = 2.6 points, which is the whole of the headroom, and it is why 89.6 and not 90. At 87% with a 10% final it is 1.3 points. A light final is a weak lever and a sharp one at the same time: little headroom, and ten points of exam for every point of course grade. A heavy final is the opposite, forgiving to aim at and dangerous to fail.
And (1 − w) ÷ w is the exam points a single point of pre-final average is worth: nine at a 10% final, four at 20%, three at 25%, one and a half at 40%.
The equation above assumes a straight weighted average with nothing else going on. Grading policies are allowed to be more generous, and often are. California State University, Long Beach states the permission in its grading policy: “This does not prohibit options, such as dropping or replacing exam/assignment grades or making other adjustments that allow for student grades to be improved.” The same policy requires that the rules be written down, instructing that “instructors shall include in the syllabus the types and sequence of assignments and basis for assigning course grades”. Read your syllabus before you trust any grade calculator, including this one.
Five quizzes, equally weighted, category worth 20% of the course: 95, 88, 72, 90, 40. The average of all five is 77.0. Drop the 40 and it is 86.25. The category moves 9.25 points, so the course moves 9.25 × 0.20 = 1.85 points, and at a 25% final that is 1.85 × 3 = 5.55 points off the score you need. One line of syllabus, five and a half points of exam.
The trap is timing. Many gradebooks apply the drop only when the term closes, so the average you are reading still contains a score that will never count.
The University of Kentucky’s general chemistry course describes the rule in its own words: “If offered in your course (see syllabus for details), the replacement option allows you to replace your lowest semester exam grade with the corresponding section of the final exam.” It is bounded and one-directional. “Only your one, lowest exam grade can be replaced”, and “It can only help your grade.” It does not ask why the score was low: “If your lowest grade is a zero due to an unexcused absence, that is the grade that will be replaced.”
The implementation is usually a drop rather than a swap. The University of Texas at Austin’s documentation for its Quest system describes the mechanism: “This will then drop the lowest number in the exam bucket (that now contains the final), achieving the end goal of functionally replacing the lowest test score with the final (if higher).” And if the final is not higher, “that is the score that will be dropped, leaving the exam scores untouched.”
What this does to the equation on this page is change the final’s weight, and by much more than people expect. Take three tests at 20% each, scored 62, 85 and 91, homework at 15% scored 95, and a final at 25%.
Without the rule, course = 0.20(62 + 85 + 91) + 0.15(95) + 0.25F, which is 61.85 + 0.25F. An 80 in the course needs 72.60 on the final.
With the rule, and assuming the final beats the 62, the final fills that test slot as well: course = 0.20(F + 85 + 91) + 14.25 + 0.25F, which is 49.45 + 0.45F. An 80 now needs 67.89.
Same course, same scores, one syllabus line
The final is not worth a quarter of that grade. It is worth 45% of it. The rule is worth 4.71 points of required exam score, and in exactly the same motion it makes a bad final nearly twice as damaging, which is the half of the rule nobody mentions.
Because the rule is conditional, the course grade is not one line but the higher of two: 61.85 + 0.25F and 49.45 + 0.45F. They cross at F = 62, where both give 77.35, which is the arithmetic agreeing with the policy: below your lowest test the replacement does nothing, above it the replacement is in force. A single linear equation cannot represent that, so a required score reported for a course with a replacement rule will be too high.
In a course graded on total points, 870 out of 1,000 is 87.0%. Add 25 points of extra credit that raise points earned and leave points possible alone, and it is 895 out of 1,000, or 89.5%. A 2.5-point lift.
In a weighted-category course the same 25 points land inside one category and are then multiplied by that category’s weight. Added to a homework category of 200 points, they take it from 170 (85.0%) to 195 (97.5%), a 12.5-point lift inside the category. If homework is 10% of the course that is 12.5 × 0.10 = 1.25 points of course grade, half what the identical 25 points were worth in the points-based course. Where extra credit is posted matters as much as how much of it there is.
A category is not necessarily the average of its percentages. The University of Pittsburgh’s Canvas guidance states the rule: “Assignments preserve their relative weight (based on points) within an assignment group. An assignment worth more points than another in the same assignment group will contribute more to that group’s value.”
Two quizzes in one category, one worth 10 points and one worth 100. You score 10 out of 10 and 50 out of 100. Totalling the points, the category is 60 of 110, or 54.55%. Averaging the two percentages, which is what almost everyone does in their head, gives (100 + 50) ÷ 2 = 75.00%. The two methods disagree by 20.45 points, and in a category weighted at 20% that is 4.09 points of course grade. Check which method your course uses before trusting an average you worked out yourself.
Whether 89.5 becomes an A− is not arithmetic. It is a policy, and the institution often declines to set it at all. Kentucky’s chemistry department answers the question directly: grade cutoffs “are established by each instructor for their course and are given in the course syllabus”, and the university “does not establish any grade cutoffs based on points or percentages”. Brigham Young University’s registrar says the same of the whole institution: “There is no official university grading scale (e.g. 92-100 A, 89-91 A-, etc.). Individual departments may have a specified scale that instructors use.”
Where a rounding policy exists, it is usually implemented as a moved threshold rather than a rounded number. The University of Oklahoma’s guidance for instructors describes it that way: “Canvas allows you to adjust grade thresholds—such as setting an A to start at 89.5% to align with your rounding practices.”
So run the calculator twice. At 87% going in with a 20% final, a target of 90 needs 102 on the exam and a target of 89.5 needs 99.5. That is the whole distance between impossible and barely possible, and it rests on one sentence in a syllabus. If the sentence is not there, plan against the hard cutoff, because the hard cutoff is the only version you can hold anyone to.
What the cutoff is varies more between institutions than most students realise: on five named published scales, the grade percentage calculator turns one score into three different letters.
Institutions regulate when a heavy assessment may be given, and the threshold is specific. Penn State’s Faculty Senate policy 44-20 instructs faculty to “Schedule any end-of-semester examinations worth more than ten percent of the course grade during the final examination period”, and to “Schedule no examinations during the last week of classes. (Quizzes and narrowly limited tests in support of classroom instruction worth no more than ten percent of the semester grade may be given during the last week of classes.)”
Ten percent is where an assessment stops being a quiz and starts being a final, as far as that policy is concerned. If your syllabus states a percentage, use it; if it states points, divide the final’s points by the total points in the course.
Your instructor’s scale, and whether anything is rounded. This page works in percentages and takes no view on what percentage is an A. The sources above put that cutoff in the instructor’s hands.
Any curve. The University of Illinois guidance for instructors separates the two approaches: grades “may be obtained by comparing a student’s performance with specified absolute standards rather than with such relative standards as the work of other students”, and under an absolute standard “It is possible in a given class that all students could receive an ‘A’ or a ‘B.’” Under a relative standard a required percentage means nothing until the class distribution is known, and the guide is not kind about the method: “Grading on the curve is efficient from an instructor point of view. Therein lies the only merit in the method.”
Whether your components really weigh what the syllabus says they do. The same Illinois guidance works through two exams intended to count equally that in practice counted about two to one, because their score spreads differed: “Because there is so little variability on the 80-item exam, if we merely add each student’s scores together, the spread of scores in the total will be very much like the spread of scores observed on the first exam.” Nothing on this page detects that, and neither does your gradebook.
Required scores at or below zero. When the target is already secured by the work behind you, the required score this page reports falls below zero. A negative percentage is not a score. Read anything at or below zero as “a zero on the final still meets this target”, then check the syllabus for an attendance or minimum-final-score rule, which this equation cannot see.
A final weight of zero. The weight box expects a number between 1 and 100. A course with no final is not something this equation can solve, because the final’s weight is the divisor, and the figure reported at or near zero weight is not meaningful.
Anything outside this one course. This page is the arithmetic inside a single course. Grade points across several courses are a different calculation: see the GPA calculator and the course average calculator.
The gradebook of record. The grade that counts is the one your instructor submits to the registrar. This page has no access to it, and where the two disagree, this page is the one that is wrong.
Every quotation above was read against its named source on 1 October 2026. Nothing you type leaves your browser. More: the MyCampusKit tools.
It depends on two numbers, and the arithmetic is (target − current × (1 − weight)) ÷ weight. At 87% going in with a final worth 20%, a 90 needs 102 on the exam, which means it is already out of reach: the best that course can finish at is 89.6. At 87% with a final worth 40%, the same 90 needs 94.5, which is hard but real. The final’s weight matters more than the gap does.
That the target is unreachable, not that it is difficult. Work out the ceiling instead: current × (1 − weight) + 100 × weight. At 78% with a final worth 15% that is 81.3, so the highest grade still available is 81.3 and nothing in the exam period changes it. Only three things lift a ceiling: extra credit that actually exists, a drop or replacement rule in your syllabus, or a correction to a mark already entered.
Usually not without adjusting it. Ohio State’s Canvas guidance describes the default figure as adding up “the graded assignments only”, and says that “Ungraded assignments are ignored in this calculation”. This page needs your average across the whole non-final portion, so if two of ten homeworks are unmarked, put your own estimate in for them. At a 20% final, being one point out on that average moves the required score by four points.
No, and it overstates what you need. With three tests at 20% each scored 62, 85 and 91, homework at 15% scored 95, and a final at 25%, an 80 in the course needs 72.60 on the final without the rule and 67.89 with it, because the rule raises the final’s effective weight from 25% to 45%. It cuts both ways: the same rule makes a poor final almost twice as costly.
By more than people expect. Five quizzes of 95, 88, 72, 90 and 40 average 77.0; drop the 40 and they average 86.25. In a category worth 20% of the course that is 1.85 points of course grade, and at a 25% final it is 5.55 points off the score you need.
Nobody can tell you from outside your syllabus. The University of Kentucky’s chemistry department states that cutoffs “are established by each instructor for their course and are given in the course syllabus” and that the university “does not establish any grade cutoffs based on points or percentages”, and BYU’s registrar says “There is no official university grading scale”. It is worth running both numbers: at 87% with a 20% final, a 90 needs 102 and an 89.5 needs 99.5.
Your syllabus decides, but policy sets a floor on what counts as one. Penn State’s Faculty Senate policy 44-20 tells faculty to schedule any end-of-semester examination “worth more than ten percent of the course grade” in the final examination period, and allows only quizzes “worth no more than ten percent of the semester grade” in the last week of classes.
No. This is one weighted-average equation with no knowledge of your syllabus, your instructor’s scale, any curve, extra credit, late penalties or attendance rules. The grade of record is the one your instructor submits.