A science fair abstract is a summary of 250 words or fewer that states your purpose, your procedure, your key results with numbers, and a one or two sentence conclusion, written in the past tense after the experiment is done. Below are eight complete examples across fields, a fill-in template, the official ISEF rules for 2027, and the mistakes judges notice first.
Still choosing a project? Start with science fair project ideas. No prediction written yet? Use the science fair hypothesis examples first, because the abstract reports whether that hypothesis held up.
What a science fair abstract is
The UNC Writing Center defines an abstract as a self-contained, short statement that describes a larger work. A science fair abstract is the informative kind: it reports results and a conclusion, not only the topic. Society for Science says ISEF judges use the abstract to decide whether a project merits special awards, so it is often your first impression.
| Part | Length | What it answers |
|---|---|---|
| Purpose | 1 to 2 sentences | Why does the problem matter, and what did you test? |
| Procedure | 2 to 3 sentences | What did you do, with how many trials or samples? |
| Results | 2 to 3 sentences | What were the key numbers? |
| Conclusion | 1 to 2 sentences | Was the hypothesis supported, and what does it mean? |
ISEF abstract rules for 2027
From the 2026 to 2027 International Rules and the Society for Science abstract guide. Regional fairs may use their own form.
| Rule | What it means for you |
|---|---|
| 250 words maximum, one page | The header (title, names, school, city, state, country) sits above the body. |
| Written after experimentation | Write it when the data is in, in the past tense. |
| Current year's work only | Continuation projects summarize this year, with minimal reference to past work. |
| Your work, not your mentor's | Leave out procedures done by a mentor or before you joined. |
| No acknowledgments | Do not name your mentor, lab, university, awards or patents. |
| No commercial product names or logos | Write "a 10-watt solar panel," not the brand. |
| No tables, graphs, images or bibliography | Key numbers go in sentences; references go on Form 1A. |
| Your own words | Finalist abstracts are run through a plagiarism checker, and students may not use generative AI to write them. |
At ISEF, only the official stamped Abstract and Certification may be displayed, so the rules suggest leaving an 8.5 by 11 inch space on your board. The science fair poster layout guide shows where it fits.
Science fair abstract template
Copy this, replace every bracket, and delete the labels before you submit.
[Title: what you tested, in plain words]
[Your name] | [School, City, State, Country]
PURPOSE: [One sentence on why the problem matters.]
This project tested whether [independent variable] affects [dependent variable].
PROCEDURE: [What you built, grew, surveyed or coded.]
[Groups, number of trials or participants, what you measured and how.]
RESULTS: [Key result with units, compared with the control.]
[Second result or statistical test, for example p = 0.01.]
CONCLUSION: The results [supported / did not support] the hypothesis that [...].
[One application or next step.]
How to write a science fair abstract
Free download: The Science Fair Project Kit
Seven printable worksheets that take a project from first idea to judge-ready: hypothesis builder, experiment design, data log, board layout, and the ISEF scoring rubric with interview prep.
8 science fair abstract examples by field
Each example is original and follows the ISEF rules: past tense, no brand names, no acknowledgments, numbers in sentences. The first is a full ISEF-length abstract. The rest are shorter, which suits most local and regional fairs.
Environmental engineering (full 250-word example)
Removing Microplastic Fibers from Laundry Greywater with a Sand and Biochar Filter
Microplastic fibers shed during laundry pass through household drains and enter wastewater plants that were not designed to capture them. This project tested whether adding a layer of biochar to a gravity-fed sand filter would increase the removal of microplastic fibers from simulated laundry greywater. Three filter designs were built in 50 cm PVC columns: sand only, sand with a 5 cm biochar layer, and sand with a 10 cm biochar layer. Each filter received 2 liters of greywater spiked with a known count of polyester fibers per liter. Fibers in the outflow were collected on filter paper, stained with Nile red, and counted under a fluorescence microscope. Each design was tested in 10 trials, and flow rate was recorded for every trial. The sand-only filter removed a mean of 61% of fibers. The 5 cm biochar filter removed 84%, and the 10 cm biochar filter removed 91%. A one-way ANOVA showed a significant difference between designs (p = 0.002), and a post hoc test found that both biochar filters outperformed sand alone. Flow rate fell by 18% with the 10 cm layer, which still met the target of filtering one load of laundry within an hour. The results supported the hypothesis that biochar increases microplastic capture, likely because its porous, rough surface traps fibers that slip between sand grains. A biochar layer could be added to low-cost household greywater systems, and future work will test how many loads the filter can handle before it clogs.
Plant sciences
Effect of Hydrogen Peroxide Seed Priming on Radish Germination Under Salt Stress
Soil salinity reduces germination on irrigated farmland. This project tested whether soaking radish seeds in dilute hydrogen peroxide improves germination in salty soil. Seeds were soaked for 8 hours in distilled water (control) or in 0.5%, 1% or 2% hydrogen peroxide, then planted in trays watered with 100 mM sodium chloride, with 100 seeds per group. Germination was counted daily for 7 days. Seeds primed in 1% hydrogen peroxide reached 78% germination versus 52% for the control, and their roots were 31% longer. The 2% treatment lowered germination to 44%. The results partly supported the hypothesis: a moderate dose improved salt tolerance, but the effect reversed at the highest dose.
Microbiology
How Cooking Time Changes the Antibacterial Activity of Garlic Extract
Garlic contains allicin, an antibacterial compound that heat may destroy. This study measured how heating time changes the effect of garlic extract on nonpathogenic E. coli K-12. Extract was heated at 100 degrees Celsius for 0, 5, 10 or 20 minutes, and discs soaked in each were placed on seeded agar plates, with water discs as the control and 6 plates per group. Unheated extract produced a mean inhibition zone of 14.2 mm, which shrank to 9.1 mm after 5 minutes and 4.3 mm after 10 minutes, with no zone after 20 minutes (r = -0.94). The results supported the hypothesis that heat reduces garlic's antibacterial activity, so adding garlic late in cooking may preserve it.
Chemistry
Banana Peel Powder as a Low-Cost Adsorbent for Copper in Water
Copper from mining and old plumbing can contaminate drinking water. This project tested whether dried banana peel powder removes copper ions, and how dose affects removal. Peel powder was added at 0.5, 1, 2 and 4 grams per 100 mL to 50 mg/L copper sulfate solutions, stirred for 60 minutes, filtered, and measured with a colorimeter, with 5 trials per dose. Removal rose from 38% at 0.5 g to 89% at 2 g, then leveled off at 91%. The data fit the Langmuir model (R squared = 0.97), suggesting single-layer binding. The results supported the hypothesis that peel removes copper in a dose-dependent way up to saturation, so food waste could serve as a cheap filter material.
Physics and energy
Effect of Dust Accumulation on Solar Panel Power Output
Dust lowers solar output in dry regions, but cleaning schedules are often guesswork. This project measured how much power a small panel loses as dust builds up. A 10-watt panel was placed under a fixed halogen lamp, and fine sand was sifted onto it in steps of 1 gram per square meter, up to 10 grams. Voltage and current were recorded at each step, and the test was repeated 5 times. Output fell about 4.5% per gram per square meter (R squared = 0.98), reaching a 43% loss at 10 grams. The results supported the hypothesis that power loss rises steadily with dust density, and the model can estimate when cleaning pays for itself.
Behavioral and social sciences
Effect of Background Music With Lyrics on Reading Comprehension
Lyrics may compete for the same language processing students use to read. With Institutional Review Board approval and parental consent, 60 high school volunteers read three passages of equal difficulty in random order under silence, instrumental music and music with lyrics, then took a 10-question quiz after each. Mean scores were 8.1 in silence, 7.8 with instrumental music and 6.2 with lyrics. A repeated-measures ANOVA found a significant effect of condition (p = 0.0004), with the lyrics condition differing from both others. The results supported the hypothesis that lyrics impair comprehension, while instrumental music had little effect.
Software design and machine learning
A Compact Neural Network for Offline Tomato Leaf Disease Detection on Phones
Most image recognition models are too large to run offline on a basic phone, which limits their use on farms. This project designed a compact convolutional neural network to classify tomato leaf diseases without internet. A public dataset of 16,000 labeled images across 10 classes was split into training, validation and test sets, and the model was compared with a standard large model. The compact model reached 95.4% test accuracy versus 97.1%, while being 28 times smaller and 6 times faster on a mid-range phone. The results supported the hypothesis that a model can shrink sharply with a small loss in accuracy, letting farmers diagnose disease in the field.
Biomedical engineering
A Low-Cost Wrist Sensor for Detecting Falls in Older Adults
Falls can seriously injure older adults who live alone. This project built a wrist-worn fall detector from a microcontroller and an accelerometer that flagged a fall when an acceleration spike was followed by 10 seconds of stillness. With Institutional Review Board approval, 10 adult volunteers performed 200 simulated falls onto a mat and 400 everyday movements. The device detected 188 of 200 falls (94% sensitivity) with 9 false alarms (97.8% specificity), and the stillness check cut false alarms by 71% versus a spike-only algorithm. The results supported the hypothesis that combining impact and stillness improves detection, at a parts cost under 25 dollars.
Weak vs strong abstract
Weak: "Salt is a big problem for plants. I wanted to see if hydrogen peroxide helps seeds grow. I will soak seeds and plant them. My mentor at the university lab helped me a lot."
No numbers, future tense, a named mentor, and no outcome. The plant sciences example above is the same project rewritten so every sentence gives a judge something to score.
Common science fair abstract mistakes
| Mistake | Fix |
|---|---|
| No results, only the plan | Write it after the data is in and give the main number |
| Half the words on background | One sentence of why, then what you did |
| Vague results ("it worked better") | Value, unit and comparison: "78% vs 52%" |
| Naming or thanking a mentor or lab | ISEF bans acknowledgments; remove them |
| Describing your mentor's work | Report only your own procedures |
| Brand names | Describe the item generically |
| Future tense | Past tense throughout |
| Overclaiming ("this proves") | Say "supported the hypothesis" |
| Letting an AI tool write it | ISEF rules forbid it |
Judges compare your abstract with your interview, so be ready to explain every claim out loud. The guide on how ISEF is judged covers what they score.
Science fair abstract checklist
- 250 words or fewer, past tense
- Purpose, procedure, results and conclusion all present
- Specific numbers with units, plus sample size
- No mentor, lab, award or brand names
- Only this year's work, done by you
From science fair abstract to published paper
A strong science fair abstract is also the first draft of a journal abstract, since both use purpose, methods, results and conclusion. The gap is everything around it: a literature review, methods detailed enough to repeat, proper statistics, figures, and a discussion of limitations. Most students stall there.
The YRI Fellowship closes that gap. Over 10 weeks, a student works one on one with a PhD mentor, writes a first-author paper submitted to peer-reviewed IEEE, Springer or Elsevier venues, and prepares the same project for ISEF or JSHS. 87% of Fellows publish and 73% place at competitions. To work on the writing yourself, see how to write a research abstract and how to publish a research paper in high school.
Related guides
- Science fair hypothesis examples
- How to win science fairs
- ISEF project ideas by category
- Science fair poster layout
- High school research paper examples
Frequently Asked Questions
How long is a science fair abstract?
For ISEF and its affiliated fairs, the maximum is 250 words on one page. Many local fairs accept 150 to 200 words, so check your fair's form.
Does the title count toward the 250 words?
On the ISEF form, the title, names and school sit in a separate header above the abstract body. If your fair's form says otherwise, follow it.
Should a science fair abstract include results?
Yes. ISEF asks for the key results that lead directly to your conclusions, and judges look for specific numbers.
Can I mention my mentor in the abstract?
No. ISEF bans acknowledgments, including your mentor and research institution, and the mentor's own work must be left out. Credit your mentor on the required forms and in your paper.
Can I use AI to write my science fair abstract?
Not for ISEF. The 2026 to 2027 rules say a student may not use generative AI to write the research plan, abstract or poster.
How do I turn my science fair project into a published paper?
Expand the abstract into a full manuscript with a literature review, detailed methods, statistics and a discussion, then submit it to a journal that accepts high school authors. The YRI Fellowship pairs you with a PhD mentor for 10 weeks to write a first-author paper for IEEE, Springer or Elsevier venues while you prepare for ISEF or JSHS. Applying takes 10 minutes and has no fee.
Sources
- How to Write a Regeneron ISEF Abstract (Society for Science)
- International Rules page, 2027 rules and forms (Society for Science)
- International Rules: Guidelines for Science and Engineering Fairs 2026 to 2027, full rules book (PDF)
- ISEF 2027 Sample Abstract and Certification with 22 categories (PDF)
- Abstracts (The Writing Center, University of North Carolina at Chapel Hill)