The PhD defense is not primarily a test of whether you know your research. You are the world expert on your thesis topic. It’s a test of whether you can communicate your contribution clearly and handle technical questions under pressure. Most people pass. The preparation is about passing well and not being blindsided.
This post covers the logistics, the structure, the preparation timeline, and what actually determines whether your defense goes smoothly. Most of it applies across institutions and countries, though the exact format varies (US public talk plus closed committee exam versus European public jury, for example). The principles are the same.
The Structure of a PhD Defense: Talk Plus Questions
The format varies by institution and country. It’s worth checking your department’s specific requirements, but here’s what’s typical:
In US life sciences programs, a PhD defense usually consists of two parts. First, a public seminar (open to the entire department or institution) lasting 45-60 minutes. Your committee attends, but so do faculty, postdocs, graduate students, and anyone else interested. This is public. This is where people from other labs can ask questions.
Second, a closed committee examination lasting 60-90 minutes. Your committee (usually 4-5 people) sits down with you, usually in person though sometimes now via videoconference. Your advisor is there, but the examination is led by another faculty member (the chair, who is usually not your advisor). This part is not public. The questions are more probing. This is where the committee evaluates whether you have genuine expertise in your field.
In many European systems, the format is different. A public defense with the committee acting as a jury. Opponents (external experts) are appointed to formally critique your thesis. The defense is public and more formal.
This post focuses primarily on the US format, but the preparation principles translate.
The Public Talk: This is a Job Talk for Your Community
Your public seminar should be intelligible to a broad departmental audience. This means not just your committee, but faculty from other labs, graduate students in different subfields, postdocs who know your general area but not your specific project.
Here’s a recommended structure:
Introduction (15% of time, 7-9 minutes for a 45-minute talk)
Start with why this problem matters. Not “I did experiments on this protein” but “This protein is involved in a process that, when dysregulated, causes disease. Current therapeutics targeting it have these limitations. Understanding its mechanism would enable better treatments.”
Give them context and motivation before you give them your specific project. Establish stakes.
Background (10-15% of time, 5-7 minutes)
What was known before your work? What are the key papers? What’s the current state of the field?
This doesn’t need to be comprehensive. It needs to be sufficient context that your results will land as significant.
Chapter 1 / Project 1 Results (25% of time, 11-13 minutes)
Present your findings from your major project clearly. Data slides should lead with the conclusion stated in the slide title. “Protein X localizes to mitochondria under stress conditions” is a better slide title than “Subcellular localization of protein X.”
Show figures that clearly support your conclusions. Simple, readable figures beat complex ones.
Chapter 2 / Project 2 Results (25% of time, 11-13 minutes)
Present your second major contribution. Same principle: clear conclusions, strong data, readable figures.
Synthesis and Future Directions (15% of time, 7 minutes)
What does this work mean overall? How do your projects connect? Where could the field go from here?
This is where you show you understand the significance of your own work.
Acknowledgments (slide or final minute)
Thank your advisor, committee, lab members, funding agencies, collaborators.
Every slide must have a clear message. One key point per slide. Slides should help your talk, not be the talk. You should be able to give a compelling seminar without reading any text from your slides.
Practice until you don’t need the slides as a script. If you’re reading words off the screen, your talk isn’t ready.
The Committee Examination: What It Really Is
Your committee is not trying to trick you or fail you. They’re evaluating two things:
First, do you understand your field deeply enough to be an independent scientist? Can you place your work in context? Can you talk about work outside your specific project? Can you explain why you made the choices you made?
Second, are the conclusions in your thesis supported by your data? Are you overstating findings? Are you appropriately cautious? Do you acknowledge limitations?
Common question types:
“Why did you choose this method over alternative X?”
This is asking whether you understand the trade-offs. You should be able to say something like, “Method X would have given us higher resolution, but Method Y was three-fold faster and our goal was throughput, not maximal resolution. We used both for validation.”
“What controls did you run for Y?”
This is asking whether you did good science. Be specific about controls: positive controls, negative controls, technical replicates, biological replicates. Know your numbers.
“What are the limitations of your approach?”
This is inviting you to demonstrate intellectual honesty. Every approach has limitations. Committees are looking for scientists who understand their tools’ constraints. “Our approach uses cultured cells, and we acknowledge that cell culture doesn’t perfectly recapitulate in vivo conditions. That’s why we validated findings in animal models.”
“If you could redo this, what would you do differently?”
This is asking what you’ve learned. It’s okay to say “I would [design experiment differently] to [answer question more directly].” It shows critical thinking. It’s better than claiming you’d do everything exactly the same, which no one believes.
“Where does the field go from here?”
This is asking whether you see beyond your own work. Can you articulate the next questions? Where will this lead?
Preparing for Tough Questions: Make a List of 20
Generate a list of 20 questions you hope no one asks. These are your weak points. Your unfamiliar territory. The experiments you know are borderline. The conclusions you’re least confident about.
For every one of these, write down a thoughtful answer. Not a defensive answer, a thoughtful one.
Example weak points:
-
“You only tested this in one cell type. How do you know it generalizes?” Answer: “That’s a fair limitation. We focused on [cell type] because [good reason], and we would need to test in [other cell types] to generalize. Our preliminary data in [one other system] suggests [cautious interpretation].”
-
“This finding contradicts paper X from 2020. How do you reconcile that?” Answer: “Good catch. That paper used [different approach] and found [different result]. We believe the discrepancy is due to [explanation]. We did [experiment] to test that hypothesis.”
-
“Your statistical analysis uses [approach]. Why not [standard approach]?” Answer: “Our data violated the assumptions of [standard approach] because [reason]. [Our approach] is appropriate for [type of data]. We also did [sensitivity analysis].”
Write real answers to 20 hard questions. When you’re under pressure in the exam room, you’ll have thought through these already.
For every experiment, make sure you know:
- What the control is and what it tells you
- What controls you ran
- How many replicates
- What the alternative interpretations are
- Why you believe yours
If a committee member catches you guessing, that’s worse than admitting you don’t know.
Slide Design for the Defense: Principles That Matter
One message per slide stated as a sentence in the title.
“Protein X localizes to mitochondria under oxidative stress” is better than “Localization.” It guides the audience. They know what they’re looking at before they see the data.
Figures taken directly from papers often have axis labels and legends that are too small to read from the back of the room. Check. If you can’t read it from 10 feet away, increase the font size or redraw the figure.
No slide should take more than two minutes to present comfortably. If you’re still talking about one slide after two minutes, it’s too complicated.
Data slides should lead with the conclusion. Show the data that supports it. Don’t make the audience work to extract the message.
Avoid slides that are just text. One to two sentences on a data slide is fine. A slide that’s a paragraph is a slide you should remove or turn into two slides.
Color matters. Use colors that are visible to people with color blindness (avoid red-green combinations without additional distinction). Black text on white background is safest. Avoid busy backgrounds.
Your slides are a support for your talk, not a replacement for it. If someone could understand your entire story by just reading the slides without hearing you talk, you’re relying too much on the slides. You should be adding information verbally.
Practice Schedule: Minimum Timeline for Preparation
Eight weeks out: Outline the entire talk. What are the major sections? What’s the flow? Write it out.
Six weeks out: Draft slides. Create the visual presentation. Don’t worry about perfection yet. Just get ideas on slides.
Four weeks out: First full practice run. Give yourself 45 minutes and go through the entire talk. Time yourself. Record video if possible and watch it. This is usually humbling. You talk faster than you think. You filler-word more than you notice. You repeat yourself.
Three weeks out: Practice with your lab. Tell them this is practice. Ask for specific feedback: Did the introduction set up the problem well? Were the results clear? Did I sound confident? What was confusing?
Two weeks out: Practice with someone outside your field. A colleague from a different lab. Someone in a different department. If they can’t follow the introduction or grasp why your work matters, revise it. They’re a proxy for your committee, many of whom may not work on exactly your topic.
One week out: Practice the introduction and conclusion until they’re automatic. These are the parts where you set tone and may be most nervous. If you can deliver the first two minutes and final two minutes without hesitation, you’ve won half the battle.
Day before: Go through your slides once. Not 10 times. Once. Rest. Sleep well.
Don’t over-practice. The goal is familiarity, not memorization. If you sound robotic, you sound unprepared. If you sound like you’re reading a script, you sound unprepared. Somewhere between these extremes is a practiced talk where you know the content, you’ve rehearsed the flow, but you’re still presenting it with some energy.
Day of the Defense Logistics
Arrive early (30 minutes minimum) to test AV. Does your laptop connect to the projector? Do your figures display correctly? Are colors accurate? Is the laser pointer working?
Have your slides in two places: laptop and USB drive. If you can, have them on a cloud service too. Technical failures happen.
Bring water. You’ll be talking for 45 minutes and then answering questions for 90 minutes. Your mouth gets dry.
Have your thesis printed and on the table in front of you during committee questions. You might need to reference a specific figure or result. It should be at hand.
If you don’t know the answer to a question, say so. “I don’t know, but my best guess is X because Y” is far better than confident guessing and being wrong. Committees respect intellectual honesty.
If you need a moment to think, pause. A 10-second silence while you think is completely normal and far better than rambling while you collect your thoughts.
If a question seems unfair or is based on a misunderstanding, it’s okay to clarify. “I think I may have miscommunicated that result. Let me rephrase…” This is professional and appropriate.
After the Defense: What Outcomes Look Like
Most defenses result in immediate approval. You pass. Congratulations.
Some defenses result in “contingent pass” or “pass with minor revisions.” This means the committee is satisfied with your defense but wants you to make specific changes to the written thesis. These are usually small changes: revise a paragraph, clarify a figure, add citations, revise the abstract. You make these revisions, submit them, and you’re done.
Some defenses result in “major revisions,” which means more substantial changes to the thesis. This is less common but it happens. You revise, resubmit, sometimes have another committee meeting to review the revisions. You still pass, but it takes more time.
Very rarely, a committee might not recommend passing and ask you to defend again. This is uncommon for students who’ve done reasonable preparation. If it happens, your department usually provides support to address the committee’s concerns and then you defend a second time.
Know your institution’s specific policies, but in general, most committees are invested in your success. They’ve already agreed to be on your committee, which means they believe you have done doctorable research. They want you to pass. They’re not looking for reasons to fail you.
Bottom Line
If you want a book that covers the full arc of graduate school — from choosing a program through the dissertation and beyond — Getting What You Came For by Robert Peters remains the most practical guide available. The defense preparation chapters are especially good.
Two weeks before your defense, your preparation is largely done. What matters most in the final two weeks is sleep, confidence building, and mental preparation, not cramming more practice. Do your practice now. Then trust that you’ve done the work. You know this material better than anyone in the room. Your job is to communicate that clearly. Get sleep. Take care of yourself. Walk in knowing you have this.