In short
When you see a trinomial, scan for two perfect squares — one at the start, one at the end. Take their square roots and call them \sqrt{\text{first}} and \sqrt{\text{last}}. Now check the middle: if it equals +2 \cdot \sqrt{\text{first}} \cdot \sqrt{\text{last}}, the trinomial is (\sqrt{\text{first}} + \sqrt{\text{last}})^2. If it equals -2 \cdot \sqrt{\text{first}} \cdot \sqrt{\text{last}}, it is (\sqrt{\text{first}} - \sqrt{\text{last}})^2. The middle sign is the only thing that decides between + and - inside the bracket. Five seconds, no working.
You sit down with a worksheet. The first trinomial is x^2 + 6x + 9. The second is 25y^2 - 20y + 4. The third is 9a^2 + 12a + 6. Your friend pulls out a pen, starts splitting the middle term, drawing brackets, mumbling about factor pairs. You glance, glance, glance — and you have already factored the first two and ruled out the third. Three seconds total.
That is what this article gives you. Not a calculation, but a scan. The eye learns to do what the pen used to do.
The scan, step by step
Stand back from the trinomial and ask three questions in order:
- Is the first term a perfect square? Things like x^2, 4x^2, 9a^2, 25y^2 — anything you can write as (\text{something})^2.
- Is the last term a perfect square? Numbers like 1, 4, 9, 16, 25, 36, 49, or expressions like 9b^2.
- Does the middle term equal \pm 2 \cdot \sqrt{\text{first}} \cdot \sqrt{\text{last}}? Multiply the two square roots, double the result, and compare.
If all three answer yes, you have a perfect-square trinomial. The sign of the middle term then tells you whether the bracket is a sum or a difference. Why the sign rule works: (a+b)^2 = a^2 + 2ab + b^2 has a +2ab middle, and (a-b)^2 = a^2 - 2ab + b^2 has a -2ab middle. The outer two terms are identical in both expansions — only the middle sign changes. So when you read a trinomial back, only the middle sign can tell you which identity is hiding inside.
If any answer is no, the trinomial is not a perfect-square form. It might still factor by other methods, but stop chasing (a \pm b)^2 and move on.
Why this works at all
The whole trick rests on one fact about FOIL. Multiply (a+b) by itself the long way:
The first and last terms are forced to be perfect squares — they are literally a squared and b squared. Why the outer terms must be perfect squares: when you multiply a binomial by itself, the only way to get the highest-degree term is "first times first", which is a \cdot a = a^2. The only way to get the lowest is "last times last", which is b \cdot b = b^2. There is no other route. So if a trinomial is genuinely (a+b)^2 in disguise, its outer two terms are perfect squares, no exceptions.
The same expansion with a minus sign gives:
Same outer terms. Same magnitude in the middle. Only the middle sign flipped. That is the entire reason the middle sign is the deciding bit when you read a trinomial backwards. Why the middle sign is the only signal: in (a+b)^2, both cross terms ab and ba are positive — they add to +2ab. In (a-b)^2, both cross terms are -ab and -ba — they add to -2ab. The outer squares a^2 and b^2 are positive in both cases, because squaring kills any sign. So the only place the bracket's sign survives is in the middle term.
Worked examples
x^2 + 6x + 9 — what is hiding inside?
Step 1. First term: x^2. That is (x)^2. Perfect square — tick.
Step 2. Last term: 9. That is 3^2. Perfect square — tick.
Step 3. Middle term: +6x. Predicted middle: 2 \cdot x \cdot 3 = 6x. Match, and the sign is +.
All three checks passed, middle sign is positive. The hidden identity is (a+b)^2 with a = x and b = 3:
25y^2 - 20y + 4 — minus in the middle
Step 1. First term: 25y^2 = (5y)^2. Perfect square — tick. Note the coefficient: \sqrt{25y^2} is 5y, not y. Why the coefficient matters: (5y)^2 = 25y^2, so the square root is 5y as a whole, not just y. Forgetting the 5 is the silent error that wrecks the prediction in step 3.
Step 2. Last term: 4 = 2^2. Perfect square — tick.
Step 3. Middle term: -20y. Predicted middle: 2 \cdot 5y \cdot 2 = 20y. The magnitudes match. The sign in front is -.
All three checks passed, middle sign is negative. The hidden identity is (a-b)^2 with a = 5y and b = 2:
Verify by expanding: (5y-2)^2 = 25y^2 - 20y + 4. Done.
9a^2 + 12a + 6 — does the test pass?
Step 1. First term: 9a^2 = (3a)^2. Perfect square — tick.
Step 2. Last term: 6. Is 6 a perfect square? \sqrt{6} is irrational. Cross.
The test fails at step 2. Stop. This is not a perfect-square trinomial — there is no neat (a \pm b)^2 hiding inside.
It might still factor by some other method (in this case it does not factor over the rationals at all), but the perfect-square shortcut is off the table. Why a single failed step kills the shortcut: the shortcut works only when the trinomial truly is a^2 \pm 2ab + b^2 for some clean a and b. The outer two terms must literally be squares of those clean values. If the last term is not a perfect square, no b exists that makes b^2 = 6 in nice numbers, so the structure (a \pm b)^2 cannot fit.
Move on. Try splitting the middle term, the ac method, or the quadratic formula — but do not waste any more time on perfect-square hopes.
A cricket-context anchor
Picture two batsmen, a and b, opening together. They each score their own runs (a^2 and b^2) and they also build a partnership. The partnership runs come from running between the wickets together — and there are exactly two ways to count them: a's contribution to b's runs and b's contribution to a's runs. Both ways give ab. Add them: 2ab.
If both batsmen are happy partners (the + sign), the partnership adds runs: +2ab. If they are mid-fight (the - sign — one is calling no, the other is sprinting yes), the partnership costs runs: -2ab.
The middle term of a perfect-square trinomial is exactly that partnership figure. The outer terms are the individual scores. The middle sign is the mood between them.
The cornerstone of completing the square
This pattern is the foundation of the most powerful technique in school algebra: completing the square. When you face a quadratic like x^2 + 6x + 5 = 0 and you cannot factor it cleanly, you instead build a perfect square on purpose:
You took the first two terms, asked yourself "what last term would make this a perfect square?", and answered using exactly the rule from this article: \sqrt{x^2} = x, half the middle coefficient is 3, so b = 3 and b^2 = 9. Add and subtract that 9, and the first three terms collapse into (x+3)^2.
That single move — recognising the hidden (a \pm b)^2 — is what powers the quadratic formula, the vertex form of a parabola, and a chunk of integration tricks in calculus. The whole tower stands on the five-second scan you just learned.
A quick drill
Predict in five seconds — sum, difference, or neither?
- x^2 + 10x + 25 — first x^2, last 25 = 5^2, predicted middle 2 \cdot x \cdot 5 = 10x, sign +. Sum. Factors as (x+5)^2.
- 4y^2 - 12y + 9 — first 4y^2 = (2y)^2, last 9 = 3^2, predicted middle 2 \cdot 2y \cdot 3 = 12y, sign -. Difference. Factors as (2y-3)^2.
- x^2 + 7x + 12 — first x^2, last 12 is not a perfect square. Neither. (Use sum-product splitting — it gives (x+3)(x+4).)
- 16a^2 + 24ab + 9b^2 — first 16a^2 = (4a)^2, last 9b^2 = (3b)^2, predicted middle 2 \cdot 4a \cdot 3b = 24ab, sign +. Sum. Factors as (4a+3b)^2.
- x^2 - 4x + 5 — first x^2, last 5 is not a perfect square. Neither.
If you got the right answer to each one before your pen touched paper, the scan is in your eye now. That is the whole skill — recognition before calculation.