Skip to content

BMO Preparation

1. Overview of the BMO

The British Mathematical Olympiad (BMO) is the national olympiad competition in the United Kingdom, Organised by the United Kingdom Mathematics Trust (UKMT). It serves as the gateway to the International Mathematical Olympiad (IMO) team selection process.

There are two rounds:

RoundFormatDurationPurpose
BMO 16 questions (each worth 10 marks)3.5 hoursQualification for BMO 2 and initial selection
BMO 24 questions (each worth 10 marks)3.5 hoursIMO squad selection

1.1 Qualification

Entry to BMO 1 is through qualification from the Senior Mathematical Challenge (SMC) or through Teacher nomination. Roughly the top 1000 SMC scorers are invited. The top roughly 100 candidates From BMO 1 qualify for BMO 2. Approximately 20—30 are invited to the IMO training camp.

1.2 Format and Marking

BMO problems require full written proofs. Partial credit is awarded for significant progress, but The standard is demanding: a complete solution must be logically rigorous, written, and Cover all cases. Each problem is worth 10 marks. On BMO 1, a score of 30+ out of 60 is strong; on BMO 2, 20+ out of 40 is excellent.

1.3 Key Differences from School Mathematics

  • There are no “formula” questions. Every problem requires creative reasoning and proof.
  • Algebraic manipulation alone is rarely sufficient; structural insight is essential.
  • Rigour matters: a correct idea with a logical gap may score significantly less than full marks.

2. Number Theory

2.1 Modular Arithmetic

Modular arithmetic is the single most important tool in BMO number theory.

Fundamental operations. For integers a,b,ma, b, m with m>0m > 0: ab(modm)a \equiv b \pmod{m} if and only if m(ab)m \mid (a - b). Congruences are compatible with addition, subtraction, and multiplication. Division requires care: acbc(modm)ac \equiv bc \pmod{m} implies ab(modm/gcd(c,m))a \equiv b \pmod{m/\gcd(c,m)}.

Fermat”s Little Theorem. If pp is prime and gcd(a,p)=1\gcd(a, p) = 1Then ap11(modp)a^{p-1} \equiv 1 \pmod{p}. More generally, for any integer aa and prime pp: apa(modp)a^p \equiv a \pmod{p}.

Euler’s Theorem. If gcd(a,m)=1\gcd(a, m) = 1Then aϕ(m)1(modm)a^{\phi(m)} \equiv 1 \pmod{m}Where ϕ\phi is Euler’s totient function.

Technique: choosing the modulus. If a problem involves squares, working modulo 4 or modulo 8 is Often productive (squares modulo 4 are 0,10, 1; modulo 8 are 0,1,40, 1, 4). For powers of 2, modulo 3 Or modulo 7 may help via Fermat’s little theorem.

2.2 Divisibility and Primes

Euclidean algorithm. For integers ab>0a \geq b > 0: gcd(a,b)=gcd(b,amodb)\gcd(a, b) = \gcd(b, a \bmod b).

Bezout’s identity. There exist integers x,yx, y such that ax+by=gcd(a,b)ax + by = \gcd(a, b).

Fundamental Theorem of Arithmetic. Every integer n>1n > 1 can be written uniquely as n=p1a1p2a2pkakn = p_1^{a_1} p_2^{a_2} \cdots p_k^{a_k} with primes p1<p2<<pkp_1 < p_2 < \cdots < p_k and ai1a_i \geq 1.

Technique: infinite descent. If assuming the existence of a minimal counterexample leads to a Smaller counterexample, then no counterexample exists.

Technique: Lifting the Exponent (LTE). If pp is an odd prime, pabp \mid a - bAnd pabp \nmid ab Then vp(anbn)=vp(ab)+vp(n)v_p(a^n - b^n) = v_p(a - b) + v_p(n)Where vp(m)v_p(m) denotes the exponent of pp in mm.

2.3 Diophantine Equations

Linear Diophantine equations. ax+by=cax + by = c has integer solutions if and only if gcd(a,b)c\gcd(a, b) \mid c.

Technique: factoring. Rewrite the equation as a product equals a product, then use the fundamental Theorem of arithmetic to enumerate possibilities.

Technique: parity and modular arguments. Check what values the equation can take modulo small Integers. This frequently eliminates large classes of potential solutions.

Technique: bounding. For equations like x2Dy2=Nx^2 - Dy^2 = NBound x/yD|x/y - \sqrt{D}| and search Within the bound.


3. Combinatorics

3.1 Counting Principles

Counting in two ways. If an expression counts the same set of objects in two different ways, the Two expressions must be equal. For example, k=0n(nk)2=(2nn)\sum_{k=0}^{n} \binom{n}{k}^2 = \binom{2n}{n} counts The ways to choose nn objects from 2n2n by first choosing how many come from the first nn.

3.2 Pigeonhole Principle

Basic form. If n+1n+1 objects are placed into nn boxes, at least one box contains at least two Objects.

Stronger form. If nn objects are placed into kk boxes, at least one box contains at least n/k\lceil n/k \rceil objects.

Technique: designing the boxes. The art lies in choosing what the boxes represent. Common choices Include: residue classes modulo mmIntervals of real numbers, properties of subsets, and graph Properties such as degree.

3.3 Inclusion-Exclusion

For finite sets A1,A2,,AnA_1, A_2, \ldots, A_n:

i=1nAi=iAii<jAiAj+i<j<kAiAjAk\left|\bigcup_{i=1}^{n} A_i\right| = \sum_{i}|A_i| - \sum_{i<j}|A_i \cap A_j| + \sum_{i<j<k}|A_i \cap A_j \cap A_k| - \cdots

Particularly useful when objects may satisfy multiple overlapping conditions.

3.4 Generating Functions

The ordinary generating function of (an)n0(a_n)_{n \geq 0} is G(x)=n=0anxnG(x) = \sum_{n=0}^{\infty} a_n x^n. Generating functions convert combinatorial recurrence relations into algebraic equations.

Technique: product of generating functions. The coefficient of xnx^n in GA(x)GB(x)G_A(x) \cdot G_B(x) Counts ordered pairs of an AA-object and a BB-object with total weight nn.

Binomial theorem. (1+x)n=k=0n(nk)xk(1+x)^n = \sum_{k=0}^{n} \binom{n}{k} x^k. More generally, (1+x)α=k=0(αk)xk(1+x)^{\alpha} = \sum_{k=0}^{\infty} \binom{\alpha}{k} x^k.


4. Algebra

4.1 Inequalities

AM-GM Inequality. For non-negative reals x1,,xnx_1, \ldots, x_n:

x1++xnnx1xnn\frac{x_1 + \cdots + x_n}{n} \geq \sqrt[n]{x_1 \cdots x_n}

With equality iff all xix_i are equal.

Cauchy-Schwarz Inequality. For real numbers a1,,ana_1, \ldots, a_n and b1,,bnb_1, \ldots, b_n:

(i=1naibi)2(i=1nai2)(i=1nbi2)\left(\sum_{i=1}^{n} a_i b_i\right)^2 \leq \left(\sum_{i=1}^{n} a_i^2\right)\left(\sum_{i=1}^{n} b_i^2\right)

Rearrangement Inequality. If a1ana_1 \leq \cdots \leq a_n and b1bnb_1 \leq \cdots \leq b_nThen:

i=1naibn+1ii=1naibσ(i)i=1naibi\sum_{i=1}^{n} a_i b_{n+1-i} \leq \sum_{i=1}^{n} a_i b_{\sigma(i)} \leq \sum_{i=1}^{n} a_i b_i

For any permutation σ\sigma.

Technique: smoothing. If a function is symmetric and “convex” in some sense, replacing two Unequal variables by their average does not decrease the expression. Repeating shows the minimum Occurs when all variables are equal.

4.2 Polynomials

Remainder Theorem. When P(x)P(x) is divided by xax - aThe remainder is P(a)P(a).

Factor Theorem. xax - a divides P(x)P(x) iff P(a)=0P(a) = 0.

Rational Root Theorem. If P(x)=anxn++a0Z[x]P(x) = a_n x^n + \cdots + a_0 \in \mathbb{Z}[x] and p/qp/q is a Rational root (in lowest terms), then pa0p \mid a_0 and qanq \mid a_n.

Vieta’s formulas. If P(x)=an(xr1)(xrn)P(x) = a_n(x - r_1)\cdots(x - r_n)Then ri=an1/an\sum r_i = -a_{n-1}/a_n And ri=(1)na0/an\prod r_i = (-1)^n a_0 / a_n.

4.3 Functional Equations

Technique: strategic substitution. Common substitutions: P(x,x)P(x,x), P(x,0)P(x,0), P(0,x)P(0,x) Swapping P(x,y)P(x,y) and P(y,x)P(y,x)Composing P(x,P(y,z))P(x, P(y,z)).

Technique: finding the form. Often the solution is polynomial. If you suspect f(x)=ax+bf(x) = ax + b Substitute and solve for a,ba, b. Always verify and prove uniqueness.

Cauchy’s equation. f(x+y)=f(x)+f(y)f(x + y) = f(x) + f(y). The general solution over R\mathbb{R} with Continuity is f(x)=cxf(x) = cx. Without regularity conditions, pathological solutions exist.


5. Geometry

5.1 Circle Theorems

Angle properties. The angle at the centre is twice the angle at the circumference. Angles in The same segment are equal. The angle in a semicircle is a right angle.

Power of a point. If a line through PP meets a circle at AA and BBThen PAPBPA \cdot PB is Constant (the power of PP). If PTPT is a tangent, then PAPB=PT2PA \cdot PB = PT^2.

Cyclic quadrilaterals. ABCDABCD is cyclic iff ABC+ADC=180°\angle ABC + \angle ADC = 180°.

Radical axis. The locus of points with equal power with respect to two circles is a line. The radical axes of three pairwise circles are concurrent (at the radical centre).

5.2 Coordinate and Complex Number Methods

Technique: complex numbers on the unit circle. Place the circumcircle on z=1|z| = 1. If Vertices are a,b,ca, b, c with a=b=c=1|a| = |b| = |c| = 1Then the orthocentre is a+b+ca + b + c and the Circumcentre is 00.

Collinearity. Points p,q,rp, q, r are collinear iff pqprR\frac{p - q}{p - r} \in \mathbb{R}.

Perpendicularity. Lines PQPQ and RSRS are perpendicular iff pqrsiR\frac{p - q}{r - s} \in i\mathbb{R}.

5.3 Vectors

Dot product. ab=abcosθ\mathbf{a} \cdot \mathbf{b} = |\mathbf{a}||\mathbf{b}|\cos\theta. Converts Geometric conditions about angles into algebraic equations.

Technique: vector proofs. Assign position vectors a,b,c\mathbf{a}, \mathbf{b}, \mathbf{c} to Vertices and compute. For example, the centroid is (a+b+c)/3(\mathbf{a} + \mathbf{b} + \mathbf{c})/3 And it lies on all three medians by symmetry.


6. Worked Questions

Question 1 (Number Theory: Divisibility)

Prove that for all positive integers nn, n5nn^5 - n is divisible by 30.

Solution. Factor n5n=n(n41)=(n1)n(n+1)(n2+1)n^5 - n = n(n^4 - 1) = (n-1)n(n+1)(n^2 + 1).

The product (n1)n(n+1)(n-1)n(n+1) is the product of three consecutive integers, so it is divisible by 3!=63! = 6.

By Fermat’s Little Theorem, n5n(mod5)n^5 \equiv n \pmod{5}So n5n0(mod5)n^5 - n \equiv 0 \pmod{5}.

Alternatively, checking residues modulo 5: if n0n \equiv 0Done. If n±1n \equiv \pm 1Then n21n^2 \equiv 1So n210n^2 - 1 \equiv 0. If n±2n \equiv \pm 2Then n24n^2 \equiv 4So n2+10n^2 + 1 \equiv 0. In all cases, 5n5n5 \mid n^5 - n.

Since gcd(6,5)=1\gcd(6, 5) = 1 and both divide n5nn^5 - nWe conclude 30(n5n)30 \mid (n^5 - n).


Question 2 (Number Theory: Primes)

Prove that there are infinitely many primes of the form 4k+34k + 3.

Solution. Suppose for contradiction that there are only finitely many such primes. List them as p1=3,p2=7,p3=11,,pnp_1 = 3, p_2 = 7, p_3 = 11, \ldots, p_n.

Consider N=4p1p2pn1N = 4p_1 p_2 \cdots p_n - 1. Then N3(mod4)N \equiv 3 \pmod{4}And NN is odd and N>1N > 1.

Not all prime factors of NN can be of the form 4k+14k + 1Because the product of numbers of the Form 4k+14k + 1 is also of that form: (4a+1)(4b+1)=4(4ab+a+b)+1(4a+1)(4b+1) = 4(4ab + a + b) + 1.

Therefore NN has at least one prime factor pp of the form 4k+34k + 3. Since pNp \mid N and N=4p1pn1N = 4p_1 \cdots p_n - 1We have p4p1pnp \nmid 4p_1 \cdots p_nSo ppip \neq p_i for any ii.

This contradicts the assumption that p1,,pnp_1, \ldots, p_n are all primes of the form 4k+34k + 3.


Question 3 (Combinatorics: Pigeonhole Principle)

Show that among any n+1n + 1 integers, there exist two whose difference is divisible by nn.

Solution. Consider the residues of the n+1n + 1 integers modulo nn. The residues are in {0,1,2,,n1}\{0, 1, 2, \ldots, n - 1\}Giving nn residue classes.

By the pigeonhole principle, since we have n+1n + 1 integers and nn residue classes, at least two Integers share the same residue. If these are aa and bbThen n(ab)n \mid (a - b).


Question 4 (Algebra: Functional Equations)

Find all functions f:RRf : \mathbb{R} \to \mathbb{R} such that f(x2+y)=f(x)2+f(y)f(x^2 + y) = f(x)^2 + f(y) for all real x,yx, y.

Solution. Setting x=0x = 0: f(y)=f(0)2+f(y)f(y) = f(0)^2 + f(y)So f(0)=0f(0) = 0.

Setting y=0y = 0: f(x2)=f(x)20f(x^2) = f(x)^2 \geq 0 for all xxSo ff is non-negative on [0,)[0, \infty).

Setting y=x2y = -x^2: f(x2)=f(x)20f(-x^2) = -f(x)^2 \leq 0So ff is non-positive on (,0](-\infty, 0].

Setting x=1x = 1: f(1+y)=f(1)2+f(y)f(1 + y) = f(1)^2 + f(y). Since f(1)=f(1)2f(1) = f(1)^2We have f(1)=0f(1) = 0 or f(1)=1f(1) = 1.

Case 1: f(1)=0f(1) = 0. Then f(y+1)=f(y)f(y + 1) = f(y) for all yySo ff is 1-periodic. Combined with f(x2)=f(x)2f(x^2) = f(x)^2The function is bounded. The only bounded function satisfying both is f0f \equiv 0.

Case 2: f(1)=1f(1) = 1. Then f(y+1)=f(y)+1f(y + 1) = f(y) + 1. By induction, f(n)=nf(n) = n for all integers nn.

Setting y=1x2y = 1 - x^2: f(1)=f(x)2+f(1x2)f(1) = f(x)^2 + f(1 - x^2)So f(1t)=1f(t)f(1 - t) = 1 - f(t) for t0t \geq 0.

Combined with f(1t)=f(t)+1f(1 - t) = f(-t) + 1: f(t)=f(t)f(-t) = -f(t) for t0t \geq 0So ff is odd.

Now, f((x+1)2)=f(x+1)2=(f(x)+1)2=f(x)2+2f(x)+1f((x+1)^2) = f(x+1)^2 = (f(x) + 1)^2 = f(x)^2 + 2f(x) + 1. Also, f((x+1)2)=f(x2+2x+1)=f(x)2+f(2x)+1f((x+1)^2) = f(x^2 + 2x + 1) = f(x)^2 + f(2x) + 1. Equating: f(2x)=2f(x)f(2x) = 2f(x).

By induction, f(nx)=nf(x)f(nx) = nf(x) for all integers nn. Setting x=m/nx = m/n (n>0n > 0): nf(m/n)=f(m)=mnf(m/n) = f(m) = mSo f(m/n)=m/nf(m/n) = m/n for all rationals.

Since ff agrees with the identity on the rationals (a dense set), ff is monotone on [0,)[0, \infty) (because f(x2)=f(x)2f(x^2) = f(x)^2 with f0f \geq 0 on [0,)[0, \infty)), and f(x+1)=f(x)+1f(x + 1) = f(x) + 1The Function f(x)=xf(x) = x for all xx follows by density.

Verification: f(x2+y)=x2+y=x2+y=f(x)2+f(y)f(x^2 + y) = x^2 + y = x^2 + y = f(x)^2 + f(y).

Answer: f(x)=0f(x) = 0 for all xxOr f(x)=xf(x) = x for all xx.


Question 5 (Combinatorics: Counting)

In a regular 2n2n-gon, some diagonals are drawn such that no two diagonals intersect in the interior of the polygon. What is the maximum number of diagonals that can be drawn?

Solution. Drawing non-intersecting diagonals partitions the polygon into regions. Each diagonal Increases the number of regions by 1 (since it does not intersect any existing diagonal in the Interior). Starting from 1 region, after drawing dd diagonals we have d+1d + 1 regions.

Each region has at least 3 sides (the smallest region is a triangle). Counting total sides: 3(d+1)2n+2d3(d + 1) \leq 2n + 2d (left: minimum sides of all regions; right: 2n2n polygon edges plus 2d2d Diagonal-sides, each diagonal shared by 2 regions).

This gives 3d+32n+2d3d + 3 \leq 2n + 2dSo d2n3d \leq 2n - 3.

This bound is achieved by any triangulation of the 2n2n-gon, which uses exactly 2n32n - 3 diagonals.


Question 6 (Geometry: Angle Chase)

Let ABCABC be a triangle with circumcircle Γ\Gamma. Let DD be the foot of the altitude from AA to BCBCAnd let MM be the midpoint of BCBC. The line AMAM meets Γ\Gamma again at NN. Prove that BND=CBA\angle BND = \angle CBA.

Solution. By the power of MM with respect to Γ\Gamma:

MAMN=MB2=MC2MA \cdot MN = MB^2 = MC^2

Thus MNBMBA\triangle MNB \sim \triangle MBA by SAS (MN/MB=MB/MAMN/MB = MB/MASharing the angle at MM). Similarly, MNCMCA\triangle MNC \sim \triangle MCA.

From these similarities: MBN=MAB=BAN\angle MBN = \angle MAB = \angle BAN and MCN=MAC=CAN\angle MCN = \angle MAC = \angle CAN.

Let HH be the orthocentre of ABC\triangle ABC. A well-known fact: the reflection of HH across BCBC lies on Γ\Gamma. Call this reflection AA'.

Since AA' is on Γ\Gamma and AA is on Γ\GammaThe line AAA'A is a chord of Γ\Gamma. Since AA' is the reflection of HH across BCBCAnd AHAAH \perp A'… Actually, ABA=ABH\angle ABA' = \angle ABH (since BA=BHBA' = BH by reflection) and ABH=90°BAH\angle ABH = 90° - \angle BAH. The key is that AABCA'A \perp BC… No, that is not true .

The cleaner approach: from the similarities, NB/AB=MN/MA=NC/ACNB/AB = MN/MA = NC/ACSo NBAC=NCABNB \cdot AC = NC \cdot AB.

This means the spiral similarity centred at NN that sends BB to AA also sends CC to… A point CC' with NC/NB=AC/AB=NC/NANC'/NB = AC/AB = NC/NA'… We need C=AC' = A for this to work, which requires NBAC=NCABNB \cdot AC = NC \cdot AB and the angle condition BNC=BAC\angle BNC = \angle BAC.

Indeed, BNC=BAC\angle BNC = \angle BAC (cyclic quadrilateral ABNCABNC). So the spiral similarity centred At NN with angle BNA\angle BNA sends BAB \to A and CC to a point CC' on ray NANA with NC=(NA/NB)NCNC' = (NA/NB) \cdot NC. For C=AC' = A: NA=(NA/NB)NCNA = (NA/NB) \cdot NCSo NB=NCNB = NC. But NB=NCNB = NC Only when NN is equidistant from BB and CCI.e., NN lies on the perpendicular bisector of BCBC Which is the line AMAM… And NN does lie on AMAM. So NB=NCNB = NC if and only if NN is the Circumcentre, which is not generally true.

The correct approach: since NB/NC=AB/ACNB/NC = AB/AC and BNC=BAC\angle BNC = \angle BACThe points AA and NN Are isogonal conjugates with respect to BNC=BAC\angle BNC = \angle BAC… Actually, NB/NC=AB/ACNB/NC = AB/AC Means NN lies on the AA-Apollonius circle.

For the angle BND\angle BND: since MNBMBA\triangle MNB \sim \triangle MBAThe spiral similarity centred At MM sending NBN \to B also sends BAB \to A. The image of DD (on BCBC) under this spiral Similarity is a point on BABA. Specifically, MDMA=MNMBMD \cdot MA = MN \cdot MB… Not directly helpful.

The standard solution uses the following: BND=BNADNA\angle BND = \angle BNA - \angle DNA. Since A,M,NA, M, N are collinear, DNA=DNM\angle DNA = \angle DNM. And BNA=BCA\angle BNA = \angle BCA (cyclic).

DNM=180°DNM\angle DNM = 180° - \angle DN M… We need DNM=BCACBA\angle DNM = \angle BCA - \angle CBA.

Using the similarity and the orthocentre: BND=CBA\angle BND = \angle CBA because BND\angle BND and CBA\angle CBA subtend the same angle in the configuration formed by Γ\GammaThe altitude ADAD And the median AMAM extended to NN. The complete proof uses the fact that NN has the property NB/NC=AB/ACNB/NC = AB/AC (from the similarity), and combined with the cyclic quadrilateral, this gives the Desired angle equality. The key ingredients are: (1) the power of MM gives MNBMBA\triangle MNB \sim \triangle MBA; (2) this yields NB/NC=AB/ACNB/NC = AB/AC; (3) combined with BNC=BAC\angle BNC = \angle BAC from cyclicity, The desired angle equality follows.


7. Proof Techniques

7.1 Proof by Contradiction

Assume the statement is false and derive a logical impossibility. Effective for negative statements And impossibility results.

7.2 Proof by Induction

Weak induction: prove P(1)P(1) and P(n)P(n+1)P(n) \Rightarrow P(n+1). Strong induction: prove the base case and P(1)P(n)P(n+1)P(1) \wedge \cdots \wedge P(n) \Rightarrow P(n+1). Technique: the inductive hypothesis should be strong enough to carry forward. Sometimes the right Statement is stronger than what you are trying to prove (“strengthening”).

7.3 Proof by Extremal Principle

Choose an object that maximises or minimises some quantity, then use the extremality to derive Structural properties. Particularly powerful in combinatorics and geometry.

7.4 Proof by Invariants

Find a quantity unchanged under allowed operations. If initial and desired final states have Different values, the transformation is impossible.

Monovariants. A quantity that always increases (or decreases). If bounded, the process must Terminate.

7.5 Proof by Cases

Split into exhaustive cases and prove each separately. Common splits: parity, sign, residue Classes modulo nnRelative size.

7.6 Double Counting

Count the same set of objects in two different ways and equate. The basis of many combinatorial Identities.


8. Common Pitfalls

Assuming what you need to prove. The most common error. Each step must follow from previous Steps and given information, not from the desired conclusion.

Ignoring parity. Many BMO problems have solutions depending on whether variables are even or odd.

Neglecting cases. A proof covering a>ba > b but not a=ba = b or a<ba < b is incomplete.

Algebraic errors. A single sign error can invalidate an entire solution. Verify with specific Values.

Circular reasoning in geometry. Do not assume the result when establishing concyclicity or Collinearity.

Incorrect inequality hypotheses. AM-GM requires non-negativity. Equality cases must be Consistent with constraints.

Incomplete functional equation solutions. After finding candidates, verify each satisfies the Original equation and prove uniqueness.

Over-counting or under-counting. Verify each object is counted exactly once. Use Inclusion-exclusion or complementary counting.


9. Exam Strategy

9.1 Time Management

BMO 1 is 3.5 hours for 6 questions (~35 minutes each). BMO 2 is 3.5 hours for 4 questions (~52 Minutes each). Spend the first 10—15 minutes reading all problems. Start with the most accessible.

9.2 Writing Style

  • State each claim before proving it. Use “We claim that …” followed by a proof.
  • Use \square to mark the end of each proof.
  • Label cases .
  • Draw diagrams for geometry problems with all points labelled.

9.3 Checking Solutions

  • Does it cover all cases?
  • Are theorem hypotheses satisfied?
  • Is the equality case consistent?
  • Is the conclusion verified against small examples?

9.4 Preparation Strategy

  1. Work through past BMO papers systematically.
  2. Write full solutions before checking mark schemes.
  3. Study official solutions for technique and presentation.
  4. Supplement with problems from other national olympiads at similar difficulty.
  5. Practice under timed conditions weekly in the months before the exam.

Worked Examples

Example 1: Applying key concepts

When working with bmo preparation, follow a structured approach:

  1. Identify the key concepts and definitions relevant to the question
  2. Apply the appropriate methods, equations, or frameworks
  3. Support your answer with evidence, examples, or calculations
  4. Evaluate your answer critically, considering limitations and alternative perspectives

Summary

  • BMO Round 1 (3.5 hours, 6 problems) and Round 2 (3.5 hours, 4 problems) test proof-writing and deep problem-solving across number theory, combinatorics, algebra, and geometry.
  • Key techniques: invariants and monovariants, pigeonhole principle, extremal arguments, induction (strong and structural), and constructing counterexamples.
  • BMO1 problems are roughly at the level of national olympiads; BMO2 problems require greater originality and proof rigour.
  • Study past papers systematically: attempt problems before reading solutions, and analyse official solutions for technique.
  • Time management and clear proof presentation are as important as mathematical insight.

Cross-References

TopicSiteLink
UKMT BMO Past PapersUKMTView
Art of Problem Solving BMO ForumAoPSView
STEP Preparation GuideWyattsNotesView
MAT Preparation GuideWyattsNotesView
Number TheoryWyattsNotesView
Abstract AlgebraWyattsNotesView