TL;DR Singapore O Levels are not graded on a bell curve. They use standards-referenced grading: your grade depends on how well you meet the syllabus learning outcomes, not on how you rank against other candidates. "Moderation" is a separate fairness process that adjusts for paper difficulty and marking consistency --- it is not a bell curve either. If every student masters the content, every student can get A1. There is no quota.
The myth: "O Levels are graded on a bell curve"
Walk into any Secondary 4 classroom in Singapore and you will hear some version of this claim: only a fixed percentage of students can score A1, the rest are forced into lower grades, and your fate depends on how strong your cohort is rather than how well you actually know the material.
This belief is widespread. Parents discuss it at school talks. Students repeat it in group chats. Tuition centres sometimes reinforce it, implying that the only way to secure a top grade is to outperform everyone else.
The idea is appealing because it matches what many people experienced in university, where bell-curve grading (also called norm-referenced grading) is common. In a true bell curve system, a fixed distribution is imposed --- say, only 15 % of students receive an A --- regardless of how well the class performs. If everyone does brilliantly, some students are still pushed down to fill the lower grade buckets.
But national exams in Singapore do not work this way.
What MOE actually says: standards-referenced grading
In a Parliamentary reply on 9 January 2023, the Ministry of Education stated that national examinations such as the GCE O-Level and A-Level use standards-referenced grading [1]. This means grades are determined by how well candidates demonstrate mastery of the syllabus learning outcomes, not by how candidates perform relative to one another.
In a standards-referenced system, the benchmark is the syllabus, not the cohort. If every student meets the A1 standard, every student receives A1. There is no mathematical mechanism forcing a bell-shaped distribution.
What standards-referenced grading means in practice
SEAB (the Singapore Examinations and Assessment Board) sets grade boundaries after each examination sitting. These boundaries define the minimum raw marks needed for each grade. The boundaries are informed by:
The difficulty of the paper that year, as judged by senior examiners and subject specialists.
The expected standard for each grade, based on syllabus learning outcomes.
Historical data, to ensure consistency across years.
If a paper turns out harder than intended, the grade boundaries are lowered so that a student who demonstrates the same level of understanding as an A1 candidate in a previous year still receives A1, even if their raw mark is lower. If the paper is easier, boundaries are raised.
This is not a bell curve. It is a calibration process that keeps the meaning of each grade stable from year to year.
What moderation actually is
Moderation is often confused with bell-curving, but it serves a completely different purpose. Moderation ensures consistency --- across markers, across schools, and across years.
Marker-level moderation
When hundreds of examiners mark thousands of scripts, some natural variation in marking standards is inevitable. SEAB uses moderation procedures --- including double-marking, sample checks, and statistical analysis --- to ensure that a script receiving 28/40 from one examiner would receive a similar mark from another. This is a quality-control process, not a ranking mechanism.
School-level moderation for O-Level practical marks
O-Level science practical assessments (Paper 3 for pure sciences, Paper 5 for combined sciences) are partly school-based. Teachers mark their own students' practical work during the year, and these marks feed into the final grade.
Because marking standards differ between schools --- some mark strictly, others generously --- SEAB moderates the school-assessed practical marks. The process works roughly like this:
SEAB compares a school's average practical marks against the school's average theory marks (which are externally marked and therefore standardised).
If a school's practical marks are much higher than its theory performance would predict, the practical marks are adjusted downward.
If a school's practical marks are much lower than expected, they are adjusted upward.
This protects students at strict-marking schools from being unfairly penalised, and prevents students at lenient-marking schools from receiving an unearned advantage.
This is not a bell curve. No student is being pushed down to fill a quota. The adjustment is about aligning marking standards, not imposing a distribution.
At A-Level, H2 science practical papers (Paper 4) are externally examined by SEAB-appointed examiners, not by school teachers. This means the moderation challenge is between markers rather than between schools. The practical mark goes directly into the final grade at its stated weightage (20 % for all H2 sciences), without the school-level adjustment described above.
The grading system itself remains standards-referenced --- the same principle applies. Your A-Level grade depends on your absolute performance against syllabus standards, not your ranking.
Why the bell curve myth refuses to die
If there is no bell curve, why do grade distributions look like one?
The answer is straightforward: student ability in any large population naturally distributes in a roughly bell-shaped pattern. Most students cluster around the middle, with fewer at the extremes. When you plot the number of students at each grade, the resulting chart often resembles a bell curve.
But the curve is a result, not a mechanism. SEAB does not force the grades into that shape. The distribution emerges because of how human performance naturally varies, not because of an artificial cap on top grades.
Other factors reinforce the myth:
University grading uses bell curves. Parents and older siblings who experienced norm-referenced grading at university assume national exams work the same way.
Grade boundaries are not published. Because SEAB does not release exact cut-off marks, students fill the information vacuum with speculation.
The word "moderation" sounds like "bell curve" to a non-specialist. When students hear that marks are "moderated," many assume this means marks are redistributed along a curve.
Competitive culture. In a system where students are intensely focused on relative performance (L1R5, COP), it is natural to assume that grading is also relative.
What this means for your revision strategy
Understanding how grades actually work should change how you prepare:
Focus on mastery, not ranking. Your grade depends on how well you demonstrate understanding of the syllabus learning outcomes. Work through the syllabus systematically rather than trying to guess what percentage of students will score higher than you.
Use the syllabus document as your checklist. Every assessable learning outcome is listed in the SEAB syllabus. If you can confidently address each one, you are on track for a strong grade regardless of what other students do.
Do not panic about a "strong cohort." Even if your year group is exceptionally capable, that does not reduce your chances of scoring A1. Standards-referenced grading means there is room at the top for everyone who earns it.
Prepare for the practical component seriously. Because practical marks are moderated to ensure fairness, your school's marking strictness is accounted for. Focus on building genuine practical skills --- measurement technique, data analysis, planning --- rather than worrying about whether your school marks "too strictly." If you are deciding between Pure Science and Combined Science (which affects your practical paper format and weighting), see our Combined Science vs Pure Science decision guide.
No. The GCE O-Level uses standards-referenced grading, as confirmed by MOE in a January 2023 Parliamentary reply [1]. Your grade is determined by how well you meet the syllabus learning outcomes, not by how you perform relative to other candidates. There is no fixed quota for any grade.
But my teacher said there is a bell curve. Are they wrong?
Many teachers use the term "bell curve" loosely to describe the shape of the grade distribution, which does often resemble a bell curve. But the distribution is a natural outcome of how student performance varies --- it is not imposed by SEAB. Your teacher may also be thinking of university-style grading, which does use norm-referencing. At the national exam level, the system is standards-referenced.
Why do grade boundaries change every year?
Grade boundaries are adjusted to account for differences in paper difficulty. If a paper is harder than usual, the boundary for each grade is lowered so that students are not penalised for facing a tougher exam. If the paper is easier, boundaries rise. This ensures that an A1 in 2024 represents the same level of mastery as an A1 in 2025 or any other year.
Is the bell curve different for different subjects?
There is no bell curve for any subject. All subjects at O-Level and A-Level use the same standards-referenced approach. However, the proportion of students achieving each grade does vary between subjects, because different subjects attract different student populations and have different difficulty profiles.
Does moderation mean my marks could be lowered?
Moderation of school-based practical marks can adjust marks upward or downward, depending on whether your school's internal marking is more lenient or more strict than the national standard. If your school marks strictly, moderation is more likely to raise your practical marks. If your school marks generously, moderation may lower them. The purpose is fairness, not punishment.
If there is no bell curve, why is it so hard to get A1?
A1 requires a high level of mastery across the entire syllabus. The difficulty comes from the breadth and depth of the content, not from an artificial cap on how many students can achieve the grade. The reason most students do not get A1 is that the standard is genuinely demanding --- not because SEAB limits the number of top grades.
How does this affect my L1R5 or L1R4 score?
Your L1R5 (or L1R4) aggregate is simply the sum of your individual subject grades. Since each subject grade is determined by standards-referenced assessment, your aggregate reflects your absolute performance across subjects. A strong cohort does not inflate your aggregate; a weak cohort does not deflate it.
Key takeaways
Singapore O Levels use standards-referenced grading, not a bell curve. Your grade depends on your mastery of the syllabus, not your ranking.
Moderation is a fairness adjustment that accounts for paper difficulty and marking consistency. It is not a bell curve.
There is no quota for A1 or any other grade. If every student meets the standard, every student receives the grade.
Grade distributions look bell-shaped because student performance naturally varies that way --- the curve is a result, not a mechanism.
The most effective revision strategy is to focus on meeting every syllabus learning outcome, not on trying to outperform other students.
