Titu's lemma is Cauchy-Schwarz in a shape that fits on one line, and that shape is what makes it useful. If you can see a sum of fractions with squares on top, the lemma replaces the whole sum with a single fraction. It is also called the Engel form or Sedrakyan's inequality.
The statement
For real numbers and positive reals ,
Equality holds when the ratios are all equal. The may be negative or zero. The must be positive.
The proof
For two terms, clear the denominators. The claim is
Put both sides over . The left numerator is , and the right numerator is . Subtract: what is left is . So the difference of the two sides is , which is a square over a positive number. That is non-negative, and it is zero exactly when , that is, when .
For terms, apply the two-term case again and again, merging the first two fractions, then the result with the third, and so on. Or read it as Cauchy-Schwarz applied to the vectors and . Their dot product is , so . Divide by and you have the lemma.
The signals
- A sum of fractions where each numerator is a square: . This is the direct case.
- A sum of fractions where the numerator is not a square, but you can multiply top and bottom by something to make it one. becomes .
- A sum such as . Write the numerators as , , .
- The denominators add up to something the problem controls, either a fixed constraint or an expression that simplifies after summing.
- The problem asks for a lower bound or a minimum of a sum of fractions.
Example: Nesbitt in three lines
For positive reals , prove
The numerators are not squares, so multiply each fraction by its own numerator on top and bottom. The sum becomes . Titu's lemma gives
Now , because . So the right side is at least . This is the inequality from the archive's Nesbitt problem, and there is a guide on Nesbitt's inequality that compares three ways in.
Example: IMO 1995, Problem 2
The problem: for positive reals with , prove
This looks nothing like a Titu problem. The cubes and the single power of the denominators hide it. Substitute , , , so that . The first term becomes
The other two terms follow by symmetry. Here , which is the step from the first fraction to the second. Now Titu's lemma applies directly:
Finally, by AM-GM, since . So the sum is at least .
When a problem has a product constraint and the terms have mismatched degrees, try replacing each variable with its reciprocal and see if squares appear on top.
Example: a minimum with numbers
Minimise for positive with .
Write it as . Titu gives at least . Equality needs , so , , . Plug those in: . The bound is reached, so is the minimum.
A lower bound is the minimum only if some input reaches it, and the equality condition in Titu's lemma tells you which input to try.
Where it goes wrong
- Negative denominators. With and , the left side is and the right side is . The inequality is false. Check the signs of the first.
- Applying it where the equality case is impossible. The bound is true but may not be the minimum. The ratios must be able to be equal under the constraints.
- Choosing the wrong numerators. In Nesbitt you could also write and Titu would give , which is at most and usually below it, so it does not finish. The numerators decide how the denominators combine, so try the choice that makes the new denominators simplest.
- Forgetting the last step. After Titu you usually still owe a second, simpler inequality. Write it down so you do not stop early.
What to do with this
Take every sum of fractions you have failed on and try to make the numerators squares. If you succeed, write the Titu bound and look at the new denominator. If it collapses to something known, you are done. The AM-GM and Cauchy-Schwarz technique page covers the family this belongs to, and the guide on AM-GM or Cauchy-Schwarz helps you decide between them. The problem archive has more inequalities to practise on.
Related guides
- How to prove an olympiad inequality, step by stepMost failed inequality proofs pick a tool in the first minute. The order of attack that picks it last: equality case, degree, substitution, shape, audit.
- Nesbitt's inequality: three proofs and how to spot it in disguiseNesbitt's inequality takes two lines to prove once you know it. The skill is seeing it inside a harder problem, and choosing which of three proofs to use.
- The rearrangement inequality in olympiad problems: when to use itThe rearrangement inequality says a sum of products is largest when both lists are sorted the same way. Proof by swapping, the signals, and Nesbitt by rearrangement.
- Jensen's inequality: how to set one up on a competition problemJensen is one line of theory and three lines of setup, and the setup is where it goes wrong. The shape that calls for it, and the two checks that save you.
- AM–GM or Cauchy–Schwarz: how to know which one to useThe two inequalities solve different shapes, not different difficulties. A shape test that picks one in seconds, and the equality case that settles the ties.
Techniques in this guide
- Classical InequalitiesEquality tells you where to aim.
Competitions this guide helps with
- IMOsix proof problems, three on each of two days · 4.5 hours on each of two consecutive days
Train it on Lemma
Every technique here is taught explicitly, in order. 128 lessons, 1237 curated problems and unlimited generated practice at six difficulties. Free to start, no card, and the paid plan opens with 3 free days.
Find your level