How ISEF Is Judged: The 100-Point Rubric Almost Nobody Reads

Regeneron ISEF publishes exactly how projects are scored. The rubric is public, specific, and posted by Society for Science, and in our experience almost no family reads it before entering. Students spend months perfecting the science and walk in not knowing that a third of their score has nothing to do with their data. Here is the rubric, line by line, and what each section actually rewards. The same criteria govern the regional and state fairs that feed ISEF, so this applies to your local fair too.

The 100 points

For science projects, the Grand Award rubric splits like this:

Read that last row again. Presentation is worth 35 points, more than any other section, and 25 of those points are the interview. Your poster, the thing students spend weeks perfecting, is worth 10. The conversation they rarely rehearse is worth two and a half times the board.

And you do not present once. At ISEF, each project is judged at least four times in 15-minute interviews by different judges.

Section by section

Research Question: 10 points

Judges look for a clear, focused purpose, a stated contribution to the field, and a question testable with accepted methods. "Cancer" is not a research question. "Whether a specific gut enzyme metabolizes a specific drug class" is. The single most common failure here is a topic instead of a question.

Design and Methodology: 15 points

Variables and controls defined and complete. This section rewards the boring virtues: what did you control, what did you compare against, why this method rather than the alternative. If you cannot defend the method choice, not just the method, points leave here.

Execution: 20 points

Systematic data collection, reproducibility, correct statistics, and, crucially, whether the data actually supports the conclusion drawn. Judges hunt the gap between what your results show and what your conclusion claims. Overclaiming is the most expensive mistake in this section, and it is epidemic among strong students.

Creativity and Potential Impact: 20 points

Judges explicitly weight originality and impact on the field, technology, economy, environment or society. A competent replication of a known result scores badly here no matter how well executed. This is the section that punishes safe projects.

The Engineering variant

Engineering projects are scored on a parallel rubric: Research Problem (10) looking for a practical need, solution criteria and constraints; Design and Methodology (15) covering alternatives explored and a prototype built; Execution: Construction and Testing (20) requiring the prototype tested across multiple conditions and trials; then the same Creativity (20) and Presentation (35). If you built something, know which rubric you are under, because "where is your control group" and "how many conditions did you test" are different questions.

The interview: the 25 points nobody practices

From the published criteria, judges in the interview assess: clarity of responses, understanding of the basic science behind the project, interpretation and limitations of the results, degree of independence, recognition of potential impact, and the quality of ideas for further research.

Two of those deserve special attention:

Limitations are scored. Volunteering what your study cannot show earns points. Pretending it has no weaknesses loses them, because every judge knows there is one, and the interview exists to find out whether you do.

Independence is scored. Judges explicitly assess how much of the project is yours. A student who cannot explain a choice their mentor made loses points, which is why good mentorship means understanding every decision in your own project, not outsourcing them.

The four questions that decide most interviews: What would you do differently? What is the biggest weakness in your study? Why this method and not another? What is the next experiment? Walk in with real answers to those four and you are ahead of most of the room.

What this means for your preparation time

If presentation is 35% of the score, your preparation should look something like the rubric. The students we coach split their final month roughly in half: half on the paper and poster, half on mock interviews where someone who knows the field asks the hard questions before a judge does. It feels excessive until the first judge asks question three.

This is precisely what YRI Fellows prepare for. Our students do original research 1-on-1 with full-time PhD mentors, and then we run mock judging against this exact rubric before their fairs. 73% of our Fellows place at competitions, and this rubric is a large part of why. If you want that preparation for your own project: yriscience.com.

Rubric weightings summarized from the Regeneron ISEF Grand Award judging criteria published by Society for Science. Read the original before your fair.