Wet lab vs dry lab, and how to choose a lab that fits you

Three kinds of work, one confusing word

"Lab" gets used to mean a physical room, a research group, and a type of work, and those three meanings don't line up. A computational lab has no bench. A clinical research group might work entirely out of a hospital office. Someone can say "I work in Dr. Chen's lab" and mean any of these.

The three broad categories are wet lab, dry lab, and clinical research. Wet lab means hands on biological material at a bench. Dry lab, also called computational or in silico work, means analyzing data on a computer. Clinical research means studies involving human participants, their data, or their samples, usually attached to a hospital or clinic.

Plenty of labs mix them. A genomics lab might have bench people generating sequencing data and computational people analyzing it. A cancer lab might have a clinical trials arm. This is good news, because it means picking a lane at nineteen is not a life sentence. But the day-to-day of your specific role will be firmly in one of these categories, and you should know which one you're signing up for before you say yes.

One more piece of vocabulary you'll need throughout. The PI, or principal investigator, is the faculty member who runs the lab and holds the grants. Below them are postdocs (finished their PhD, working in the lab for a few years), graduate students (working toward a PhD), technicians or research assistants (often full-time paid staff), and undergraduates. Who you actually work with matters more than any of the rest of this.

Wet lab, day to day

Wet lab is physical work with your hands on biological or chemical material. If you've pictured research, this is probably what you pictured, and it's the version with the highest romance-to-reality gap.

A real day looks like this. You arrive and thaw reagents. You spend forty minutes making a buffer solution and labeling forty tiny tubes with a pen that keeps smearing. You pipette the same three microliters into ninety-six wells, which is more tedious and more physically demanding on your hand and shoulder than anyone tells you. You start an incubation and have ninety minutes of dead time in which you can read a paper or clean up, but you can't leave. You run the gel. The gel is blank. Nobody knows why the gel is blank.

If the lab does cell culture, cells are alive and they don't observe your class schedule. They need feeding on their own timeline, including weekends. If the lab does animal work, there's mouse colony maintenance, genotyping, cage checks, and a real emotional dimension to working with animals that you should think honestly about before committing. If the lab does anything with human samples, pathogens, radioactivity, or hazardous chemicals, there's mandatory safety training before you touch anything.

The defining constraint is that you must be physically present, in contiguous blocks, often at times you don't control. An experiment with a timepoint at hour six starts when it starts. You cannot do a Western blot from your bedroom at midnight after a shift.

What's genuinely great about it: you learn to work with your hands, you develop real technical skill, you build the intuition for how biological knowledge is actually produced, and the moment an experiment works after four failures is a specific kind of satisfaction that's hard to get elsewhere. Wet lab is also the most direct preparation if you're heading toward a PhD or MD-PhD in bench science.

Dry lab, day to day

Dry lab, computational, or bioinformatics work means you and a computer. The material you work with is data that someone else generated, or data you pull from public databases.

A real day looks like this. You open a terminal. You spend two hours discovering that the sample names in one file use a different naming convention than the sample names in another file, and writing code to reconcile them. You run an analysis. It fails with an error message you don't understand. You search the error message, read documentation, try three things, and the fourth works. You make a plot. The plot looks wrong. You find the bug. You make the plot again and this time it shows something real, and that's a good day.

The work is usually Python or R, sometimes bash and command-line tools, plus statistics. Depending on the lab it might be RNA-sequencing analysis, genome-wide association studies, single-cell data, medical imaging, electronic health record data, protein structure, epidemiological modeling, or machine learning.

Two honest things about the barrier to entry. First, it's low in terms of equipment: you need a laptop and internet, not a badge and a biosafety training certificate. Many computational labs will take a motivated undergrad who has done a couple of online courses, and a lot of the work can be done remotely, at whatever hour suits you. Second, the learning curve at the start is genuinely brutal, and it feels worse if you've never coded before and everyone around you has. That feeling is normal and it does pass, usually around month three.

If you want to test whether this fits you before committing, real free resources exist. The Carpentries (carpentries.org) runs free lesson material on Python, R, and the command line built specifically for researchers. Rosalind (rosalind.info) is a free set of bioinformatics problems you solve by writing code. "R for Data Science" by Wickham and Grolemund is free online. freeCodeCamp is free. Work through a few weeks of any of these and you'll know quickly whether the frustration feels interesting or just miserable.

Clinical research, the third door

Clinical research is research involving human participants: their outcomes, their data, their samples, their experience. It's the category most pre-med students never hear about, and it's very often the most accessible one.

A real day looks like this. You pull up a list of patients who came through the emergency department in the last two years matching a set of criteria, and you go through their charts extracting variables into a REDCap database. Or you're in clinic, screening the day's schedule for patients who might be eligible for a study, then approaching them in the waiting room to explain what the study involves and ask if they're interested. Or you're calling patients at their three-month follow-up to walk them through a questionnaire. Or you're helping a coordinator prepare an amendment to an IRB protocol, which is the document the institutional review board approves before any human-subjects research can legally happen.

Why it's accessible: the bottleneck in clinical research is almost never technical skill. It's finding people who are reliable, careful with details, comfortable talking to strangers, and willing to do meticulous work. If you can do those things, you're immediately useful. Bilingual students, especially Spanish-speaking students, are in genuine demand for consenting and follow-up work.

There's real paperwork before you start. You'll almost certainly need CITI Program training in human subjects research, which is an online course most institutions require and pay for, plus HIPAA training, and often a background check and immunization records. Budget two to four weeks for onboarding.

Two more things worth knowing. Clinical research often produces a poster, an abstract, or authorship faster than wet lab does, because retrospective chart-review studies move on a timescale of months rather than years. And it puts you in a hospital, around clinicians, watching how medicine actually gets practiced. If you're pre-med and unsure whether you want medicine at all, this answers that question better than a bench does.

The tradeoff nobody mentions if you work or commute

Most research advice is written for a student who lives on campus, doesn't have a job, and has fifteen unclaimed hours a week. If that's not you, the standard advice will make you feel like you're failing at something you're actually just structurally excluded from. So let's be direct about the scheduling reality.

Wet lab is the least forgiving. It needs contiguous blocks of at least three to four hours to be worth anything, it needs you physically in the building, and the schedule is partly set by the experiment rather than by you. If you work evening shifts, if you commute ninety minutes on public transit, or if you have caregiving responsibilities at home, wet lab is not impossible but it will cost you the most. Be honest with a PI about your constraints up front. A good one will work with you or tell you it won't work, and both answers are useful.

Dry lab is the most forgiving. The work travels. You can run analyses from a library, a kitchen table, or a laptop at 10pm after a closing shift. Meetings still happen on a schedule, and you should be present for lab meeting, but the bulk of the work is asynchronous. If your life is fragmented into odd hours, this is often the honest answer.

Clinical research sits in between and varies a lot by role. Chart review and data entry can frequently be done remotely once you have secure system access, which makes it as flexible as dry lab. Patient recruitment and consenting require you to be on site during clinic hours, which conflicts with class, but these roles are often organized into defined shifts you sign up for, which is easier to plan around than an experiment that decides when it ends.

None of this means wet lab is off the table if you work. It means you should walk in knowing which constraint you're negotiating, and you should say it out loud in the first conversation. "I work Thursday and Friday evenings and I can't move that" is a normal thing to tell a PI. What isn't normal, and what damages your reputation, is committing to hours you can't keep and then disappearing.

What each path builds, and where it leads

Think about this in terms of skills you'll still have in five years, not in terms of which one sounds most impressive on an application. All three are legitimate research experience. Admissions committees care about depth, what you understood, and whether you can talk about it, not about whether you held a pipette.

• Wet lab builds: technical hands, experimental design intuition, troubleshooting under uncertainty, patience, and an understanding of where biological data actually comes from and how noisy it is. Leads toward: PhD and MD-PhD programs in bench science, industry research and biotech roles, and any career where you need to judge whether an experiment was done well.
• Dry lab builds: programming, statistics, data cleaning, reproducibility habits, and the ability to work with datasets far larger than any one lab could generate. Leads toward: bioinformatics, computational biology, biostatistics, epidemiology, health data science, clinical informatics, and industry roles. Worth naming plainly: these are also the most transferable and most immediately employable skills of the three, which matters if you need to earn money between college and medical school.
• Clinical research builds: understanding of study design and IRB ethics, patient communication, meticulous data handling, and firsthand exposure to clinical practice. Leads toward: medical school (where it's directly relevant), clinical trials work, public health, health policy, and clinical research coordinator jobs, which are real full-time positions that many people hold in a gap year.
• All three build the same underlying thing: the ability to ask a question, gather evidence, and say honestly what the evidence does and doesn't show. That's the actual point, and it's what an interviewer is probing when they ask about your research.

Mentorship structure matters more than the research topic

Here's the advice most students get too late. The specific topic of the lab matters much less than who trains you and how much attention you get. A student who spends two years in a lab studying something they find mildly interesting, with a postdoc who teaches them carefully, will come out far ahead of a student who joined their dream-topic lab and got ignored for two years.

In most labs, the PI is not the person who trains you. Your day-to-day mentor is usually a postdoc, a senior graduate student, or a staff scientist. That person's temperament, workload, and willingness to explain things is the single biggest factor in your experience. Find out who it will be before you commit, and if possible, meet them.

Lab size shapes this in predictable ways. A large lab of twenty-five people usually has more resources, more equipment, more established protocols, and more people who might have time for you, but the PI may meet you twice a year and may not know your name for months. A small lab of four or five often means direct, regular contact with the PI, which is fantastic for letters of recommendation and for learning to think like a scientist, but it can be more fragile if a grant doesn't get renewed.

Ask directly how often the PI meets with people, and how often they meet with undergraduates specifically. Weekly one-on-ones, even fifteen minutes, is a strong signal. "I'll check in when your mentor tells me you have data" is a much weaker one. Also ask whether undergraduates attend and present at lab meeting. Labs where undergrads present treat undergrads as scientists in training. Labs where undergrads never present usually treat them as hands.

Culture is harder to read from outside but you can sense it. Do people eat lunch together or does everyone have headphones on? Does anyone laugh? When you asked a basic question during your visit, did the person answer it or make you feel stupid for asking? Trust that read. You will spend hundreds of hours in this room.

Questions to ask before you say yes

Ask these in your first conversation or your trial visit. Asking them does not make you look demanding. It makes you look like someone who takes the commitment seriously, and PIs notice that. Pick the eight or so that matter most to you rather than firing off all of them.

• Who will I work with day to day, and how much time do they realistically have for training someone?
• What would my first project be? Is it my own piece of something, or am I assisting on someone else's work?
• How many hours a week do you expect, and how flexible are they around my class schedule and my job?
• Is this position paid, work-study eligible, for course credit, or unpaid? If it's unpaid now, is there a path to paid, and what does that path look like?
• What does a successful first semester look like to you? How will I know if it's going well?
• How often do you meet one-on-one with undergraduates in the lab?
• Do undergraduates present at lab meeting? Do they go to conferences, and does the lab cover the cost?
• Are undergraduates ever authors on papers here? What does someone have to do to get there?
• What happened to the last two undergraduates who worked here, and where did they go?
• What training or certifications do I need before I start, how long do they take, and who pays for them?
• Is there summer funding, or programs you nominate students for?
• What happens when someone in the lab makes a mistake?
• Would you be willing to write me a letter of recommendation after a year, if my work merits it?
• And then, separately and privately: ask a current undergraduate or graduate student in the lab what it's actually like. Ask them over coffee, not in front of the PI. That answer is the true one, and it's the most valuable information you will get.

Red flags, and how to leave gracefully

Not every lab is a good place. Some are disorganized, some are exploitative, and a few are genuinely harmful. Being new and grateful for the chance makes it hard to see this clearly, so know the warning signs in advance.

Watch for these:

1. 1To leave well, start by choosing a natural boundary: the end of a semester, the completion of a project phase, or the start of summer. Leaving mid-experiment causes real damage and people remember it.
2. 2Tell your direct mentor first, in person or on a call, before you send any email. Do not let them find out from the PI or from a group message.
3. 3Keep the reason brief and non-accusatory. "I've decided to step back at the end of the semester so I can focus on [classes / a clinical position / my work schedule]" is complete. You do not owe anyone a full accounting of your frustrations, and airing them on the way out rarely helps you.
4. 4Give notice with real runway, two to four weeks at minimum, and offer to finish or hand off what you started.
5. 5Leave your work findable. Organize your data, write up the protocol as you ran it, label and inventory your samples, and put your files somewhere your mentor can actually access. This is the single thing that determines whether people speak well of you afterward.
6. 6Send a short thank-you email after your last day. It creates a written record, it's genuinely kind, and it keeps the door open.
7. 7Ask about a letter of recommendation while you're still fresh in their memory, even if you won't need it for two years. "Would you be willing to write me a letter in the future?" is a fair question if you did good work.
8. 8Do not ghost. Stopping replies and never coming back is the one exit that follows you, because departments are small and people talk. Even a bad lab experience should end with a clear, professional goodbye.

• Nobody can tell you who your day-to-day mentor will be, or the answer keeps changing.
• "We'll figure out a project for you later" is still the answer after a month or two.
• High turnover. Undergraduates cycle out every semester, or multiple graduate students have left the program early. Ask why. If the answer is evasive, believe the pattern.
• You're doing dishes, autoclaving, and stocking tips indefinitely with no timeline and no path to a project. Some of this is normal and expected for the first few weeks; everyone does it. A year of it with no movement is not training, it's free labor.
• Unpaid with no plan and no acknowledgment that this is a burden, in a lab that clearly has funding. Ask about work-study, hourly positions, course credit, or summer fellowships. A PI who won't even discuss it is telling you something.
• Vague expectations paired with sudden disappointment. If you're never told what success looks like but you're regularly told you've fallen short, that's a management problem, not a you problem.
• You're asked to run procedures you haven't been trained for, to skip safety steps, or to record data in a way that feels dishonest. Leave over this one, and report it if it involves safety or research integrity. Your institution has a research integrity office.
• Questions are treated as an annoyance. In a healthy lab, an undergraduate asking a basic question is a normal Tuesday.
• Comments about your background, your accent, whether you "seem like a research person," or whether you're a "culture fit." You do not have to tolerate this to get a letter.
• Pressure to work hours that will visibly damage your GPA or cost you the job you need to pay rent. A mentor who doesn't respect that you have a life outside the lab is not a good mentor.