Why rounding comes up so often
Most tests don't divide evenly into 100. On a 15-question test each question is worth 6.666…%, so 13/15 is 86.666…%. On a 7-question quiz 6/7 is 85.714…%. Gradebooks have to display these as something, and whatever they display becomes the number students argue about.
The arguments cluster at the boundaries. A student at 89.5% wants to know why they aren't at 90%. A student at 89.4% wants to know why the student at 89.5% got rounded up and they didn't. Neither question is unreasonable, and both can be answered if you decide your rule before the end of the term.
The three common rounding rules
1. Standard rounding to the nearest whole percent. 89.5% becomes 90%; 89.49% becomes 89%. This is how most people expect rounding to work and it's what many gradebooks do by default when set to show whole numbers.
2. No rounding (truncating). Only the whole number counts: 89.99% is an 89%. It's simple and strict. Students sometimes see it as unfair, because a student who is 0.01% from an A doesn't get it.
3. Rounding to one decimal place, then comparing. 89.46% becomes 89.5%, which may then be compared to an 89.5% threshold. This adds a second rounding step and is the source of a lot of confusion (see below). Most teachers are better off avoiding it.
The most important rule: round once, at the end
Rounding each assignment before averaging can change the final result. Here's a student with three tests of equal weight:
| Test | Exact | Rounded |
|---|---|---|
| Test 1 (26/29) | 89.66% | 90% |
| Test 2 (35/39) | 89.74% | 90% |
| Test 3 (31/35) | 88.57% | 89% |
| Average | 89.32% → 89% | 89.67% → 90% |
Rounding each test first adds a third of a point and moves the student from a B+ to an A-. Rounding only the final average is more accurate, and it's the approach most gradebook software uses internally. Keep full precision (at least two decimal places) on individual scores and round only the number that becomes the grade.
What about borderline grades?
Even with a clear rule, there will always be a student at 89.4%. Some teachers allow a small, published "professional judgement" band at each boundary: a student within half a point of the next grade can be moved up if their recent work shows they've reached that level. Used this way, the decision is based on evidence, not on who asks.
If you do this, apply it to every student near a boundary, not only to the students who email you. The students who never ask are often the ones who most deserve a second look.
Writing a rounding policy
A good rounding policy fits in one or two sentences on the syllabus. For example:
"Final grades are calculated from exact scores and rounded once to the nearest whole percent (89.5% and above rounds up to 90%). Individual assignments are not rounded."
That's it. The point isn't which rule you choose, it's that students know the rule before they see their grade and that everyone in the class is treated the same.
Rounding in our calculators
Every calculator on this site keeps full precision and shows two decimal places by default. You can change the display to one decimal or a whole number in the grader settings. Letter grades are always decided from the exact percentage rounded to two decimals, so 89.996% shows as 90% at two decimals and earns an A−, while 89.99% stays a B+.
Quick answers
Does 89.5 round up to an A?
Under standard rounding to the nearest whole percent, yes: 89.5% rounds to 90%, which is an A- on most scales. If your school truncates, 89.5% stays 89%, a B+. Check your school's policy.
Should I round each assignment grade?
No. Keep exact scores for each assignment and round only the final average. Rounding each assignment first can push the final grade up or down by a fraction of a point, enough to change a letter grade.
This guide explains common practice and the math behind it. Grading rules differ between schools, so always check your school or district policy. Spotted an error? Let us know and we'll fix it.