Eligibility first, essays second

This is not legal advice and it is not immigration advice. Eligibility rules, tax treatment, and what any given status permits vary by programme and by individual circumstance, and they change. Verify every requirement directly with the programme, and verify anything touching your status with your school's international student office or designated school official, or with a qualified immigration attorney, before you apply for, accept, or rely on any award.

Where eligibility actually breaks

Read the eligibility section before the prize amount. The overwhelming majority of scholarship databases are built for domestic students, and the clauses that disqualify international applicants are usually one line buried below the description.

Institutional funding is where international students win most often. Departmental awards, college-level scholarships, endowed funds tied to a major, and research or teaching assistantships are administered by your own university, are frequently silent on citizenship, and have a fraction of the applicant volume of national competitions.

The decision tree

Run every candidate through this before you read the essay prompt. It takes about ninety seconds per programme and it is the single highest-leverage habit in the whole process, because everything downstream is wasted if the answer at the top is no.

1. Does it require US citizenship or permanent residency? yes → STOP. Log the reason. Never revisit this one. no → continue 2. Does it require FAFSA or a federal aid determination? yes → ask the programme whether an institutional alternative form exists. Some have one. Most do not. no → continue 3. Does it require state residency for tuition purposes? yes → almost always a stop. Read the exact definition once before you accept that. no → continue 4. Is the award a payment for services - assistantship, paid fellowship, internship stipend, prize for work performed? yes → this may be employment. Do not apply, accept, or start work until your DSO or an immigration attorney has confirmed it for your specific status. no → continue 5. Is it restricted to a school, department, major, or community you actually belong to? yes → PRIORITISE. Small pools are where the wins are. no → continue 6. Is it non-cash - compute, credits, software, hardware, travel paid directly to a vendor? yes → PRIORITISE. Rarely citizenship-restricted, almost nobody applies. no → normal priority. Queue it by deadline.

Step four is the one that matters most and the one most likely to be answered wrongly by a forum post. Whether something counts as employment, and whether your status permits it, is a specific determination about your specific case — it is not a general fact about scholarships, and the person who can answer it correctly works at your university and answers for free.

The funding types are not interchangeable

People use these words loosely, and the differences change both who is eligible and how you should apply.

• Scholarship — money toward education, usually merit or need based, paid to you or applied to tuition. Most citizenship clauses live here.
• Grant — money for a specific purpose or project rather than for you generally. Often need-based when governmental, but project grants are judged on the project, which is a much better fit if your work is strong.
• Fellowship — funding attached to a programme, cohort, or research position. Frequently includes a stipend and sometimes constitutes employment, which is exactly where immigration status becomes a real constraint rather than a checkbox.
• Assistantship — paid work as a research or teaching assistant. This is employment; whether you can hold it depends on your status and the work authorisation rules that apply to you.
• Startup credits and non-cash awards — cloud compute, API credits, software licences, hardware. Rarely citizenship-restricted because they are not payments, frequently generous, and almost nobody applies because they do not feel like "funding".
• Competitions and hackathon prizes — prize money is treated differently from employment income in many contexts, but not in all of them. Verify before you accept.

Treat the entire final column as a list of questions for someone qualified rather than as a set of answers. Tax treatment in particular differs by award type, by treaty, and by individual circumstance, and getting it wrong is the kind of mistake that surfaces a year later.

The category most consistently underused by international students is non-cash credit programmes. Free compute, credits, and tooling do not carry the citizenship clauses that block cash awards, and for anyone building projects they remove a real constraint immediately.

Where to find them

Build the list from sources rather than from a single search. Ranked roughly by yield per hour spent for an international undergraduate:

1. Your own university's financial aid site, filtered to non-federal awards. The highest-yield source by a wide margin, and the one most students never open.
2. Your department's own page. Departmental and endowed awards often have single-digit applicant counts and are frequently silent on citizenship.
3. Your college or school within the university, which is a separate set of funds from the department and the institution.
4. The international student office itself. They often maintain a curated list of awards that international students can actually hold, which is exactly the filter you want.
5. Professional societies in your major. Student memberships are cheap or free and most run awards, travel grants, and competitions.
6. Company and foundation programmes attached to a technology, a community, or an industry. Read the eligibility line first, always.
7. Community, heritage, language, and regional funds, including ones in your home country for students studying abroad — a category almost nobody searches.
8. Startup and developer credit programmes, which are rolling year-round and have no season.
9. Conference and travel grants, which are the easiest funding to win if you have a paper or a poster, and which the undergraduate research guide covers how to get.
10. General scholarship aggregators, last. High volume, low signal, mostly built for domestic students, and the source everyone starts with.

The twelve-month rhythm

Funding has a season, and most of it is not when students look. The pattern below is a typical US academic-year shape; every programme sets its own dates, so verify each one rather than trusting the month.

The February to March institutional window is the one most consistently missed, because it sits in the middle of a term when nobody is thinking about funding and it is advertised almost exclusively on a university page nobody visits. Put it in a calendar in September.

Building the pipeline

Funding applications fail on logistics far more than on merit. Deadlines are scattered across the year, each application wants slightly different materials, and recommendation letters take weeks. Treat it as a pipeline with a tracker, exactly like a job search.

1. Build the list first, apply second. Spend one session finding thirty candidates before you write a single essay.
2. Filter on eligibility immediately and record the reason for every rejection so you never re-read the same programme in three months.
3. Record the deadline type. Hard-deadline and rolling programmes need completely different behaviour: rolling ones reward applying early, hard ones reward planning backwards.
4. Work backwards from each deadline — letters requested four weeks out, transcripts two weeks out, essay drafted two weeks out, submission three days early. Portals fail on the final day, predictably.
5. Batch by material type. Write all the 500-word essays in one sitting, request all letters in one sitting. Context switching is what makes this feel enormous.
6. Track status and outcome so next year starts from a working list rather than from zero.

The tracker schema

One row per programme, and the columns exist because each one prevents a specific failure. A spreadsheet is entirely sufficient; the schema matters more than the tool.

id stable key, so notes survive a rename programme official name url the eligibility page, not the marketing page amount number, for sorting by expected value deadline ISO date deadline_type hard | rolling | not-yet-open eligibility eligible | blocked | unknown blocked_reason the exact clause, quoted - this is what stops you re-reading it next year materials essay | transcript | letters | portfolio | form letters_needed count letters_status not-asked | asked | received essay_variant which fitted opening or closing you used submitted_on ISO date, blank until it is genuinely sent outcome pending | rejected | won | no-response outcome_date ISO date notes anything a future you would want Three views do all the work: · deadline within 21 days AND letters_status != received · eligibility = unknown AND older than 7 days · deadline_type = rolling AND submitted_on is blank

The blocked_reason column is the one people leave out and the one that saves the most time. Without it you will re-read the same disqualifying paragraph three times a year, every year, and feel each time like you might have missed something.

The arithmetic of volume

Award rates are low and mostly outside your control, which makes the strategy question quantitative rather than emotional. The figures below are illustrative — nobody publishes a reliable per-application rate and outcomes are not truly independent — but the shape of the answer holds under any reasonable numbers.

Assume a 4% success rate per application, outcomes independent. Probability of at least one win after n applications = 1 - 0.96ⁿ n = 5 1 - 0.815 = 18% n = 12 1 - 0.613 = 39% n = 25 1 - 0.360 = 64% n = 40 1 - 0.195 = 80% Now raise the per-application rate to 8% by applying only where you fit the stated criteria, in small pools, with a named contact: n = 12 1 - 0.92¹² = 1 - 0.368 = 63% Twelve well-fitted applications ≈ twenty-five poorly-fitted ones.

That last line is the whole strategy in one comparison. Fit is worth roughly twice what volume is worth, which is why the institutional and departmental awards — small pools where you meet the stated criteria exactly — outrank the national competitions you will never hear back from. But fit is capped by how many programmes genuinely match you, and volume is not, so the correct plan is to exhaust the well-fitted list first and then add volume rather than choosing between them.

Time cost, once the reusable core exists: fitted opening and closing 30 min form filling and uploads 30 min chasing transcripts, letters 30 min -------- per application 90 min 40 applications = 60 hours over a year ≈ 70 min/week The first application costs far more than 90 minutes, because that is where the core essay gets written. Every one after that is paying down a cost you already covered.

Recommendation letters

• Ask four weeks ahead, minimum. Two weeks is an imposition and it shows in the letter.
• Ask people who have seen you work, not people with impressive titles who taught you once.
• Ask one person to write for several programmes at once, and give them the full list up front.
• Send a reminder a week out and a thank-you afterwards with the outcome. You will need them again.
• If someone hesitates, give them an exit. A lukewarm letter is worse than a missing one.

What goes in the packet

Send everything below in one message with the request. It is not presumptuous — it is the difference between a letter written in twenty minutes from your material and one written in twenty minutes from memory.

1. The programme name, one line on what it funds, and a link to the eligibility and criteria page.
2. The exact deadline, and the exact submission mechanism — a portal link, an email address, or a note that they will be emailed a form.
3. Your deadline, set one to two weeks before the real one, stated as such.
4. Your current CV, as an attachment, not a link.
5. Your draft essay or personal statement, even if it is rough.
6. A reminder of the context in which they know you: which course, which term, which project, and what you did in it.
7. Three specific things you would like mentioned, each tied to something they personally witnessed.
8. Any required form, rating scale, or competency list the programme uses, filled in as far as you can.
9. One sentence making it easy to decline, and a genuine thank-you regardless of the answer.

Reusing one essay without it reading as generic

You cannot write twelve original essays and you should not try. The workable method is a core narrative plus a fitted opening and closing — the middle is genuinely reusable because it is your actual story, and the parts that make it feel tailored are short and cheap to rewrite.

REUSABLE CORE (~70% — write once, edit rarely) · The specific problem or moment that started this · What you built, studied, or did about it — with a concrete detail · What it cost and what you learned that you did not expect · Where it is going next FITTED OPENING (~15% — rewrite per application) · Names the programme's stated priority in its own words · Connects it to the first line of your core in one sentence FITTED CLOSING (~15% — rewrite per application) · What this specific funding changes about what happens next · A concrete plan, not gratitude

The same core, fitted to two different programmes

The programmes below are invented to make the mechanics visible. The core paragraph — the middle 70% — is identical in both, and is not reproduced twice.

Programme A: a departmental award whose stated criteria are "academic excellence and service to the department".

Opening: "The department's award asks about service alongside excellence, and the two have been the same activity for me. I have spent two years tutoring in the introductory sequence, and the reason I stayed is the reason I started building the thing described below." Closing: "This award would cover the two shifts a week I currently work off campus, which is the only reason I cap tutoring at four hours. I would put those hours into the Tuesday session, which is the one that consistently overflows."

Programme B: a foundation grant whose stated priority is "expanding access to computing for learners who have historically been excluded".

Opening: "Exclusion from computing rarely announces itself. It looks like a capable student going quiet in week four, and the project below started because I watched that happen to people who could program perfectly well." Closing: "The grant would fund one pilot cohort — materials, a stipend for two trained tutors, and the evaluation instrument — at a school that has already said yes. Without it the pilot happens a year later, at a smaller size, without measurement."

Two rules make the difference between reuse and laziness. First, the closing must be specific about what the money changes — both closings above name a concrete thing that happens differently, with a number in it, and neither says the word "dream". Second, the concrete detail in the core has to be genuinely concrete. A number, a name, a moment. Generic essays are not generic because they were reused; they are generic because they contain nothing that could only have been written by one person.

• Mirror the programme's own stated values in their own words, once, in the opening. More than once reads as pandering.
• Answer the question that was actually asked. A large share of applications are eliminated for answering a different prompt.
• Respect the word limit exactly. It is the cheapest possible way to lose.
• Cut every sentence that would be true of any applicant.
• Have someone who does not know your work read it and tell you what you do. If they cannot, the essay is not doing its job.

One more time on eligibility

Some awards carry conditions that interact with immigration status — how a stipend is classified, whether an assistantship counts as employment, what a taxable prize means for you, and whether accepting something now affects an application later. These are not questions to resolve by reading a forum, and they are not questions this page can answer. Your school's international student office answers them for free and answers them for your specific case; for anything consequential, an immigration attorney answers them with the weight of professional responsibility behind it.

The mechanics of work authorisation itself are in the CPT vs OPT guide and the one-year rule guide, the backward-planned timeline is in the Visa Navigator, and if a funded position turns out to be an internship rather than an award, the internship guide and the sponsorship verification guide cover what to check before you apply. The pipeline discipline that makes any of this work is the same as Apply OS.

Tools referenced in this guide

• Apply OS — deadline radar and pipeline tracking, free and browser-only.
• Visa Navigator — backward-planned work authorisation timeline.
• CPT vs OPT guide — the work authorisation mechanics behind funded positions.