Oxford and Cambridge Biology Interviews

Interviews for 2027 entry fall in December 2026 — Cambridge between the 7th and the 18th, Oxford across the month and online over Microsoft Teams.

No written test at Oxford means the interview carries everything

Oxford Biology applicants have no pre-interview paper to demonstrate ability on, so two conversations in December do the whole job. There is no other place for a strong candidate to show what they can do.

With no paper to demonstrate ability on at Oxford, the reasoning has to be built for speech: the free Biology sample shows you what that reads like on the page, and the three Biology packs — Biological Sciences, Evolution & Behaviour and Physiology, £180 each — carry worked model answers across cell biology, evolution, physiology and ecology, which serves the Cambridge panel's habit of crossing into chemistry and physics too.

The question

"Sickle-cell disease is caused by a single mutation in the gene for haemoglobin. A person who inherits two copies of the mutant gene develops the disease, which is frequently severe and historically often fatal in childhood without treatment. A person who inherits only one copy is a healthy carrier. In parts of the world where malaria has long been common, the mutant gene is carried by a far larger share of the population than in regions without malaria -- in places, well over one in ten people. How can a mutation this harmful become common in a population, rather than being eliminated by natural selection?"

Hints

Is natural selection acting on the mutant gene in isolation, or on the different combinations a person can actually carry?

There are three groups to consider, not two: no copies, one copy, two copies. Does each group face the same risk from malaria?

What, specifically, happens inside a red blood cell carrying one mutant copy when the malaria parasite tries to grow inside it?

A first attempt that does not survive: "The gene must be dominant, and having even one copy must be an advantage somehow, which is why it spread." This does not survive contact with the biology. Sickle-cell disease itself is recessive -- it takes two copies to cause the disease, and a single copy does not produce it on its own. Nor is "an advantage somehow" an answer; the question asked how, specifically, and a mechanism has to be named, or the explanation has just relabelled the mystery as a mutation.

The full answer: The full answer has to track three genotypes, not two, because each meets a malaria-endemic environment differently. Someone with two normal copies of the gene is fully susceptible to severe malaria, with no genetic protection at all. Someone with two mutant copies has sickle-cell disease, which carries its own serious mortality risk independent of malaria entirely. Someone with exactly one copy of each -- a carrier, with sickle-cell trait -- is generally healthy, and their red blood cells sickle only slightly, mostly under the low-oxygen conditions found in small blood vessels and body tissues. That slight sickling is the actual mechanism of protection: when the malaria parasite infects one of these cells and starts consuming the haemoglobin inside it, the infected cell is more likely to sickle than an uninfected one, and a sickled cell is recognised and removed by the spleen before the parasite living inside it can complete its reproductive cycle. Carriers therefore suffer far less severe malaria than people with two normal copies, while also avoiding sickle-cell disease itself, which only strikes those with two mutant copies. Natural selection, on this view, is not acting on the gene as simply good or bad -- it is acting on three genotypes with three different survival outcomes in a specific environment, and the carrier genotype comes out ahead of both of the other two where malaria is common. That is heterozygote advantage. Selection does not push the mutant gene's frequency down to zero, because carriers survive malaria better than people with two normal copies. It does not push the frequency up to fixation either, because on average a quarter of the children of two carriers will inherit two mutant copies and develop sickle-cell disease. The gene settles at whatever frequency balances the malaria protection carriers gain against the sickle-cell disease cost paid by that unlucky quarter, which is also why the frequency tracks malaria risk so closely from region to region. In a population with no malaria exposure at all, the protective benefit disappears but the disease risk does not, which is exactly the calculation you would expect to shift the balance the other way over enough generations.

This is not a hypothesis invented for the interview room; it is one of the oldest tested examples of balancing selection in human genetics. J. B. S. Haldane proposed in 1949 that an inherited blood disorder might be persisting precisely because it offered some protection against a widespread infectious disease, and in 1954 Anthony Allison published field data from East Africa showing that carriers of the sickle-cell gene suffered measurably less severe malaria than non-carriers living in the same malaria-exposed area. That is the kind of evidence the reasoning above actually needs to stop being a plausible story and start being a supported explanation: not just that the gene is common where malaria is common, which could have several other causes, but that carriers specifically survive malaria infection better than non-carriers in a shared environment. A candidate who reaches the three-genotype structure unprompted has done the hard part; naming what evidence would separate this explanation from a coincidental correlation is the step that turns a good guess into a claim someone else could go and check.

That is the level of detail every pack question gets

Nine of the thirty packs across the site come with a free sample to read before you buy; Biology is one of them. Biological Sciences runs ten questions across 47 pages, Physiology ten across 31, and Evolution & Behaviour the same hints-then-answer format across 25 -- £180 each, with 10% off two packs and 20% off three at checkout.

Question 1: Natural selection edge case

"Peacocks carry an enormous, brightly coloured tail that takes real energy to grow and maintain, and makes them easier for predators to spot and catch. If natural selection favours traits that improve survival, how can a trait like this have evolved?"

Weak answer: "Peacocks evolved big tails because female peahens like them, so the males with the biggest tails get to mate more and pass on their genes. It's natural selection because the trait helps them reproduce."

Strong answer: "There's a real tension here between natural selection, which should favour survival, and what's actually happening, which looks like it works against survival. The resolution is that evolutionary fitness isn't the same as survival -- it's reproductive success. If a costly tail reliably signals a male's genetic quality or condition, because only a genuinely fit male can afford to grow and carry such a costly ornament without being caught by predators, then females that choose tail-bearing males are, on average, choosing higher-quality genes for their offspring, even though each individual peacock's own survival odds fall. That's the logic behind Zahavi's handicap principle. There's also a self-reinforcing dynamic, Fisherian runaway selection, where a female preference for long tails and the tail-length trait can co-evolve and amplify each other once the preference exists, independent of any direct fitness signal, until it's checked by the survival cost. So natural selection isn't overridden here; it's in tension with sexual selection, and the trait persists at whatever point the reproductive benefit of being chosen balances the survival cost of carrying it. I'd want to know whether tail length correlates with any measurable health or parasite-resistance marker, to distinguish the honest-signal explanation from pure runaway selection."

Why this works: the strong answer names the precise conceptual move -- fitness is not survival -- rather than restating the question's framing, cites two specific, correctly labelled evolutionary mechanisms, states the equilibrium logic explicitly (benefit balances cost), and closes by proposing evidence that would distinguish between competing hypotheses. That last step is exactly the "propose a hypothesis and examine its own weaknesses" habit tutors are trained to reward.

Question 2: Cell signalling

"A single cell contains, in its own DNA, all of the genetic instructions needed to survive independently. Why then do multicellular organisms depend so heavily on cell-to-cell signalling instead of letting each cell act autonomously?"

Weak answer: "Cells need to communicate so the body works properly. Signalling lets hormones and other chemicals tell cells what to do, like insulin telling cells to take up glucose."

Strong answer: "The issue is a coordination problem. Every cell in a multicellular organism carries the full genome and is, individually, capable of proliferating -- that's exactly what happens in cancer, when a cell stops responding to the external signals that would normally restrain its behaviour. For a multicellular organism to function, individual cells have to suppress their own default behaviour of growth and division in favour of a specialised role that serves the organism as a whole, and that suppression can only be coordinated externally, because no single cell has information about the state of the whole organism. Signalling pathways, receptor-ligand binding triggering intracellular cascades, let a small number of extracellular signals be rapidly amplified and interpreted differently depending on which receptors and downstream machinery a given cell type expresses, so the same hormone, insulin for example, can produce different but coordinated responses in liver, muscle and fat cells. That's more efficient than hard-wiring every possible response permanently into each cell's genome, because it lets the organism reallocate resources dynamically rather than through slow genetic change. So signalling is the mechanism that converts a population of genetically identical, individually capable cells into a coordinated organism, and its breakdown, loss of responsiveness to growth-suppressing signals, is a defining step in cancer."

Why this works: the strong answer reasons from a first-principles problem (coordination among identical, capable units) rather than listing an example, supplies specific mechanistic vocabulary (receptor-ligand binding, signal transduction, tissue-specific response), and volunteers an unprompted connection to disease biology -- the kind of lateral link that shows genuine understanding rather than recall.

Question 3: Ecological trade-off

"If photosynthesis is a tree's main energy source, and taller trees intercept more light, why don't trees simply keep growing as tall as physically possible?"

Weak answer: "Trees stop growing taller because they run out of nutrients in the soil, or because it takes too long to grow that tall, so shorter trees have more time to reproduce."

Strong answer: "The limiting factor is mostly hydraulic rather than nutritional. Water is pulled up from the roots to the leaves against gravity by negative pressure generated as water evaporates from the leaves -- the cohesion-tension mechanism. As a tree gets taller, that column of water has to be held under increasing tension, and at some height the tension becomes great enough that air bubbles start to form in the xylem vessels, a process called cavitation, which blocks water transport to the tissue above the blockage. Leaves at the top of very tall trees also tend to be smaller and photosynthesise less efficiently than lower leaves, even with more light available, because it becomes progressively harder to keep them adequately hydrated. There's a cost trade-off too: taller trees have to invest proportionally more carbon in non-photosynthetic structural tissue, trunk and support wood, just to remain upright, which is energy that isn't going into growth or reproduction. Redwoods sit close to the theoretical hydraulic limit for tree height, estimated at around 122-130 metres, which is a case where a physical constraint, not a genetic or nutrient one, sets the ceiling. So the answer isn't that trees choose to stop -- it's that the marginal photosynthetic gain from extra height is eventually outweighed by the hydraulic cost of supplying water to it."

Why this works: the strong answer rejects the first plausible-sounding explanation, identifies the actual physical constraint with correct terminology (cohesion-tension, cavitation), reasons in marginal cost-versus-benefit terms rather than absolute terms, and grounds the abstract argument in a concrete, named, approximately quantified example.

Treating the interview like an exam is the mistake, and it is a reflex

Two years of training to retrieve the right answer do not switch off because a page tells you to reason instead. Under pressure, candidates reach for recall, and the reaching is visible.

Breaking that reflex takes practice against a question you have not seen before, ideally with someone pushing back on the answer you give. The free Biology sample gives you one open-ended problem to try that way, and the £180 Biological Sciences pack sets a weak answer beside a strong one on questions like the peacock's tail, so you can see which one is reciting and which one is reasoning.