Use the visiting-student rules, not a Caltech student’s shortcut.
Caltech SURF involves a research project developed with a mentor, rather than a central application that automatically supplies a laboratory. For an external student, securing a suitable mentor is therefore a meaningful part of the work. A high grade-point average alone does not resolve project fit or available supervision.
| Question | Answer or unresolved item |
|---|---|
| Who is this guide for? | Continuing undergraduates applying from outside Caltech. |
| GPA | Visiting applicants: at least 2.5/4.0. Caltech’s 2.0 threshold is not the visiting threshold. |
| Academic standing | Returning for undergraduate study; required completed terms and good standing apply. High-school dual enrollment alone is not eligible. |
| International students | Caltech-campus route may accept visiting international students; JPL restricts this route to U.S. citizens or permanent residents. |
| Mentor | A Caltech faculty mentor must approve and support the application before submission. |
| Housing | Limited, not guaranteed and not included in the SURF award. |
The following distinctions were checked September 16, 2026. Read the exact program route as well: Caltech-campus SURF and work at JPL do not have identical nationality rules.
Keep the announced 2027 cycle separate from old award figures.
The application page explicitly announces SURF 2027 opening on November 1 and lists March 1 for applications and references. In that cycle context, this guide interprets those as November 1, 2026 and March 1, 2027; the page does not supply an exact cutoff time. Confirm the dates in the opened portal before relying on that interpretation.
The main SURF page still labels its $8,110 award as a 2026 amount. Do not advertise it as the 2027 award or treat it as money in addition to free housing. The program describes ten weeks of research and written and oral outputs, not an unrestricted summer grant.
Use two dates in your own plan: a personal mentor-outreach target and the published application deadline. The first is yours to change. The second is controlled by the program. A student who begins looking for a mentor near the submission cutoff may discover that the real bottleneck is agreement on a project, not the time needed to upload a transcript.
Make a narrow, informed request.
Choose a few laboratories whose recent work you can describe in plain language. Read a current project description and enough of a relevant paper or talk to understand its question. You do not need to pretend to understand every technique. You do need a reason that this laboratory, rather than any Caltech laboratory, fits your preparation.
Mentor-outreach planning fields, not a submission-ready research proposal
A useful first note identifies your institution, academic stage, relevant preparation and one concrete research interest. Ask whether the faculty member is considering a visiting SURF student and whether there may be a suitable project. Do not ask them to invent your entire application, and do not attach a long generic proposal before knowing whether the direction interests them.
Caltech itself warns against broad, spam-like messages and suggests using genuine academic connections. A professor at your home institution may know someone whose work fits, but a connection is not permission to skip the program’s approval process. Keep a small outreach record so you do not send duplicate messages or confuse interest with a firm commitment.
Sources[2]
Know which work must be yours and which reference is separate.
The application requires a student-written proposal, two confidential recommendations including at least one faculty member, and an unofficial transcript for visiting applicants. The prospective Caltech mentor provides a separate evaluation rather than serving as one of those recommendation letters. Where you have prior research with a different mentor, the instructions identify that mentor as a recommender.
Before drafting, discuss a project that can produce a meaningful result in the available period. Write a private outline naming the question, proposed approach, evidence needed, likely obstacle and fallback. The fallback should still answer a useful part of the question. It should not be a claim that any inconclusive result automatically fulfills every goal.
Caltech’s policy says generative-AI use that is not specifically permitted is not allowed for the covered application and research materials. Ask the mentor or SFP office before using it. This guide provides planning questions, not permission to submit generated proposal text. Write the actual application under the program’s rules and verify every cited work yourself.
Sources[2]
Evaluate supervision and housing before committing.
Ask what a normal research week would involve, who reviews your work and how often you can discuss difficulties. A project title cannot answer those questions. You want a task that is challenging but feasible at your starting level and a supervision arrangement that makes learning possible.
Build a separate housing plan. Find out which options are actually available, what a reservation costs and when any deposit becomes nonrefundable. Compare that timing with the award decision and payment schedule. Do not assume that acceptance automatically secures a room or that an advertised award covers every cost.
Caltech’s FAQ reports a success figure specifically for Caltech applicants. That is not a visiting-student acceptance rate, and this guide does not use it to estimate your chances. Judge the application by fit, preparation and mentor support, then keep alternative plans alive until you have a confirmed arrangement. The useful outcome is a credible research opportunity you can complete, not a probability borrowed from a different applicant population.
Data and worked inputs
Sources[3]