What Professors Look for in a Grad Student, From the Other Side of the Desk
Asked on the help desk, 2014: “What kinds of students professors are looking for?”
On this page
- They are hiring, and hiring is expensive
- Evidence that you finish things
- Independence that can grow
- Technical fit for the next two years
- Letters that compare you to other people
- Clear writing
- Temperament, which nobody writes down
- What different professors need
- How to make these traits visible
- What matters less than you think
“I have a 3.8, two internships, and a co-authored paper. Three professors told me they’re ‘not taking students.’ What are they actually looking for?”
We got some version of that question every winter. The honest answer starts with an uncomfortable shift in perspective. A professor considering a PhD student is not admitting a candidate. They are hiring a junior colleague for five or six years, paid from money they had to win, to work on problems they have promised a funding agency they will solve.
Once you see it that way, most of what professors look for stops being mysterious.
They are hiring, and hiring is expensive
A US PhD student in a lab-based field typically costs a grant tens of thousands of dollars a year once you add stipend, tuition remission, and fees. Over five years, that is one of the largest commitments a faculty member makes. Some of the cost can shift to teaching assistantships or fellowships, but the professor is still betting time, lab space, and their reputation on you. That is also why an applicant who arrives with outside money gets a warmer reading. An NSF Graduate Research Fellowship, for example, pays a $37,000 stipend plus a $16,000 cost-of-education allowance to the institution per fellowship year, for three years of support within a five-year period, which takes most of that bill off the grant.
So “not taking students” often has nothing to do with your file. It can mean the grant didn’t come through, the lab is full, or the professor is going up for tenure and cannot absorb a new person this year. Before writing to anyone, check whether they have current funding. Our grant search is the shortcut.
When the money is there, here is what the professor is trying to see.
Evidence that you finish things
The single most persuasive item in a file is a completed piece of work. A senior thesis that was defended. A poster that was presented. A dataset that was cleaned, documented, and handed off. A paper, even a middle-author one, where you can explain your contribution.
Research is mostly the long, dull stretch between a good idea and a finished product. Professors have watched bright students stall in year three because they had never carried anything to the end. A finished project, however small, is evidence you can get through the dull stretch.
Independence that can grow
Nobody expects a first-year student to run a project alone. Faculty do want signs that you can work on a problem without being told every step: that you read the manual before asking, tried two things before reporting failure, and noticed when a result looked wrong.
In letters, this shows up as phrases like “she identified the problem with our protocol before I did” or “he taught himself the analysis pipeline.” In interviews, it shows up when you describe something that went wrong and what you did next. Have one of those stories ready, told plainly.
Technical fit for the next two years
Here is something applicants rarely think about. A professor with a new grant has a list of tasks that need doing in the next couple of years. If the grant involves field sampling and R, a student who has done field sampling and knows R can contribute in the first semester. If it involves electrophysiology, a student who has handled tissue and run rigs is easier to bring up to speed than one who has only done computation.
You do not need every skill. You do need some overlap, and you need to make it visible. Read the abstract of the professor’s most recent award on NIH RePORTER or NSF Award Search. It lists, in plain terms, what the lab promised to do. Then say in your email or statement which piece of that you already have practice with. The archived NSF award on reinforcement learning for cyber-physical systems is a good model: two sentences of its abstract tell you exactly which programming and control skills a student on that project would use.
Letters that compare you to other people
Faculty read letters closely, and they read them for comparison. “Hardworking and pleasant” is a weak letter. “One of the three strongest undergraduates I have supervised in fifteen years” is a strong one. A letter from someone who watched you do research is worth more than one from a famous name who taught you in a lecture of 300.
This is why the best preparation for grad school applications often happens a year or two earlier, in choosing whose lab to work in and making sure that person sees your work. Our page on recommendation letters from top professors covers how to ask, and whom.
Most application portals also ask whether you waive your right to see the letters. Under the federal FERPA rules, a waiver is valid only if it is in writing and the school does not require it as a condition of admission, and even after waiving you can ask for the names of the people who wrote. In the files we handled, nearly everyone waived, and the faculty we worked with gave a confidential letter more weight than one the applicant could read.
Clear writing
Every stage of a research career runs on writing: proposals, papers, progress reports, fellowship applications. A professor reading your email and statement is quietly asking whether they will spend five years rewriting your drafts sentence by sentence.
Clear writing does not mean fancy writing. It means a reader can tell what you did, why, and what happened, without rereading. Short paragraphs. Specific nouns. No padding.
Temperament, which nobody writes down
Faculty rarely say this in public, but they talk about it constantly in private. The traits that make a lab work:
- Reliability. You show up, meet the deadlines you agreed to, and tell people early when something is slipping.
- How you take criticism. Your advisor will mark up your work for years. Students who get defensive wear everyone out. Students who say “got it” and return a better draft are gold.
- Honesty about results. A student who reports a failed experiment the same day is far more valuable than one who hides it for a month.
- Getting along with the lab. You will share equipment, space, and sometimes authorship. Current grad students are often asked for their opinion after a visit weekend, so treat them as colleagues.
What different professors need
Not every faculty member is looking for the same student. Career stage changes the picture.
| Professor’s situation | What they often need | What it means for you |
|---|---|---|
| New assistant professor, first grant | Builders who can set up methods and work closely with them | Lots of attention, more uncertainty, early authorship chances |
| Senior professor, large lab | Students who can work within a structure supervised by postdocs | Strong network and resources, less direct contact |
| Professor near retirement | Often not taking new students | Ask directly before investing effort |
One early-career signal is worth knowing by name. An NSF CAREER award goes only to faculty in a tenure-track assistant professor position or equivalent, and carries a minimum of $400,000 over five years. A new assistant professor holding one has the attention of the first row of the table and much less of its funding uncertainty.
Ask current students in each lab how often they meet with their advisor. That one answer tells you more about fit than any ranking.
How to make these traits visible
None of this helps if the professor cannot see it in your file. Each trait has a natural home. Finished work belongs in the first lines of your statement and your CV, with dates and outcomes. Independence belongs in one short story, in the statement or an interview, about a problem you solved without being told how. Technical fit belongs in your email, tied to a specific aim of the professor’s grant. Temperament is the one you cannot claim yourself, so it has to come from letters, which is why you should give your letter writers a short reminder of the projects you did with them and the problems you handled. A recommender who remembers the week you rebuilt a broken pipeline will write about it. One who only remembers your grade will write about the grade.
What matters less than you think
Prestige of your undergraduate school helps at the margins but does not substitute for research experience. A perfect GPA with no research tends to get a polite shortlist and no faculty sponsor. A long list of clubs, awards, and leadership roles that have nothing to do with research is mostly skipped. Our page on how professors weigh GPA and GRE scores goes further into how numbers are read.
And enthusiasm, on its own, does very little. Every applicant says they are passionate. The professor is looking for the evidence that would make that word unnecessary.
If you want to know what a specific professor is looking for, read the most recent grant abstract and the “people” page of their lab. The abstract tells you what work is coming. The people page tells you who is graduating and which roles will be empty. Between the two, you can usually guess what kind of student they need next, and write to them as that person.
More questions on this
- Do professors prefer students who already know exactly what they want to research?
- They prefer students with a clear direction, not a fixed thesis. Showing that you understand a research question and can explain why it interests you matters more than having a dissertation planned. Rigid plans that do not fit the lab's funded work can actually count against you.
- Is it better to work with a young assistant professor or a famous senior professor?
- Each has trade-offs. A new assistant professor often gives more direct attention and needs students to build the lab, but funding and tenure are less certain. A senior professor usually has stable funding and a strong network, but you may be supervised mostly by postdocs.
- I don't have any publications. Can I still be a strong PhD applicant?
- Yes. Many successful applicants to US PhD programs have no first-author papers. Faculty look for sustained research experience, a completed project such as a thesis or poster, and letters from people who supervised that work. A publication helps, but it is not the bar.
- What soft skills matter most to a research advisor?
- Reliability and how you respond to criticism come up again and again. An advisor will spend five or more years giving you feedback on drafts and experiments. Students who take critique calmly and come back with a revised version are the ones labs want to keep.