Square roots by prime factorisation

Every perfect square is built from prime factors that come in matching pairs — and that single fact turns "find the square root" into a mechanical, foolproof procedure. Break the number down into primes, pair up the identical ones, and pick one factor from each pair: their product is the square root. No guessing, no trial multiplication. This is one of two square-root methods prescribed for ICSE Class 8 (the other is long division), and it works cleanly for any perfect square you can factorise by hand — the syllabus caps this at numbers with at most four digits, so the factor tree stays short. The method also doubles as a test: factorise fully, and if even one prime is left without a partner, the number is not a perfect square at all. Below: why pairing works, a fully worked square root, a second worked example on finding the smallest multiplier that makes a number a perfect square, and the traps that cost the most marks.

The core insight

Because squaring a number doubles every exponent in its prime factorisation, a perfect square's primes always occur in matching pairs — so factorising, pairing, and taking one factor from each pair recovers the square root exactly.

The core idea

The square root of a number xx, written x\sqrt{x}, is the non-negative number that gives xx when multiplied by itself:

x=n  ⟺  n2=x,n≥0.\sqrt{x} = n \iff n^2 = x, \quad n \ge 0.

where: xx is the given perfect square and nn is its square root.

Prime factorisation method, in four moves. First, express xx as a product of primes using a factor tree or repeated division. Second, group the identical primes into pairs. Third, confirm every prime factor pairs up completely — a perfect square never leaves a prime stranded. Fourth, take one factor from each pair and multiply those together; that product is x\sqrt{x}. In symbols, if every prime in the factorisation appears to an even power,

x=p12a×p22b×p32c×⋯⟹x=p1a×p2b×p3c×⋯x = p_1^{2a}\times p_2^{2b}\times p_3^{2c}\times\cdots \quad\Longrightarrow\quad \sqrt{x} = p_1^{a}\times p_2^{b}\times p_3^{c}\times\cdots

where: p1,p2,p3,…p_1, p_2, p_3, \ldots are the distinct prime factors of xx, and 2a,2b,2c,…2a, 2b, 2c, \ldots are their (even) exponents in xx.

Figure
A factor tree for 1296 splitting down to prime leaves which are then grouped into four matching pairs of two twos and two threes
1296 factorised into primes, then grouped into pairs 1296 16 81 2 x 2 2 x 2 3 x 3 3 x 3 Four complete pairs, none left over — 1296 is a perfect square. One factor from each pair: 2 x 2 x 3 x 3 = 36 = sqrt(1296)

Why the pairing works

Since the topic is a reversal of squaring rather than a formula to derive from scratch, the reasoning is the "derivation" here — but it is short and purely structural. Understanding why grouping primes into pairs correctly recovers the square root matters more than memorising the four steps, because it is what lets you catch mistakes and explain the method rather than just execute it, and it is exactly what an exam question testing "explain why" is checking for.

Step 1. Start from what squaring does to a factorisation. Suppose n=p1a×p2b×p3c×⋯n = p_1^{a}\times p_2^{b}\times p_3^{c}\times\cdots is the prime factorisation of some whole number nn. Squaring nn multiplies the whole expression by itself, and by the laws of exponents that doubles every single exponent: n2=p12a×p22b×p32c×⋯n^2 = p_1^{2a}\times p_2^{2b}\times p_3^{2c}\times\cdots.

Step 2. Notice what this guarantees. Every exponent in n2n^2's factorisation is even, because it is twice a whole number. So every perfect square's prime factorisation has each prime appearing an even number of times — equivalently, each prime factor can be split into pairs with nothing left over.

Step 3. Reverse the process. Given a perfect square xx, if we factorise it fully and group identical primes into pairs, every pair of pp's in xx's factorisation corresponds to exactly one pp in nn's factorisation — because pairing undoes the doubling that squaring caused. Taking one factor from each pair therefore reconstructs n=xn = \sqrt{x} exactly, prime by prime.

Step 4. See why the test direction also works. If, after full factorisation, some prime is left unpaired, its exponent in xx is odd — which is impossible for a perfect square, since Step 2 showed every exponent must be even. An unpaired prime is therefore proof that xx is not a perfect square, which is why this method also serves as a perfect-square test.

Worked example

Find 1296\sqrt{1296} by prime factorisation, then use the same number to find the smallest multiplier that turns 14701470 into a perfect square.

Step 1. Factorise 12961296 into primes, dividing by the smallest prime repeatedly.

1296=2×648=2×2×324=2×2×2×162=2×2×2×2×81=24×341296 = 2\times648 = 2\times2\times324 = 2\times2\times2\times162 = 2\times2\times2\times2\times81 = 2^4\times3^4

so 1296=2×2×2×2×3×3×3×31296 = 2\times2\times2\times2\times3\times3\times3\times3.

Step 2. Group into pairs: 1296=(2×2)×(2×2)×(3×3)×(3×3)1296 = (2\times2)\times(2\times2)\times(3\times3)\times(3\times3) — four complete pairs, nothing left over, confirming 12961296 is a perfect square.

Step 3. Take one factor from each pair and multiply: 1296=2×2×3×3=4×9=36\sqrt{1296} = 2\times2\times3\times3 = 4\times9 = \mathbf{36}. Check: 362=129636^2 = 1296. ✓

Step 4. Now apply the same pairing test to a number that is not yet a perfect square. Factorise 1470=2×3×5×7×71470 = 2\times3\times5\times7\times7. Here 22, 33 and 55 each appear only once (unpaired), while 7×77\times7 is already a complete pair. To make every prime pair up we need one more 22, one more 33 and one more 55, so the smallest multiplier is 2×3×5=302\times3\times5 = 30.

Step 5. Multiplying gives 1470×30=44100=(2×2)×(3×3)×(5×5)×(7×7)1470\times30 = 44100 = (2\times2)\times(3\times3)\times(5\times5)\times(7\times7), and taking one factor from each pair, 44100=2×3×5×7=210\sqrt{44100} = 2\times3\times5\times7 = \mathbf{210}. Check: 2102=44100210^2 = 44100. ✓

Common traps

  • Stopping the factor tree too early — leaving a factor like 1212 un-broken instead of continuing to 2×2×32\times2\times3 hides whether the primes actually pair up.
  • Taking two factors from a pair (writing 2×22\times2 instead of just 22) when building the final square root, instead of the correct one factor per pair.
  • Assuming a number is a perfect square just because it looks factorisable — always check that every prime is paired with none left over before concluding it is one.
  • Forgetting to multiply the "one from each pair" values together at the end, and listing the primes separately as if that were the answer.
  • Mis-copying a repeated prime factor as two unrelated numbers — for example writing 49=7×749 = 7\times7 but then treating the two 77s as separate primes when pairing, which throws off the whole count.

Ask the tutor: Stuck factorising a number into primes, or not sure why an unpaired prime rules out a perfect square — ask the AI tutor to walk through your number step by step.

Why do we only take one factor from each pair instead of both?

Because the pair itself represents the squaring that already happened. If nn contains one factor of 33, then n2=n×nn^2 = n \times n contains two factors of 33 — that is the pair you see in xx's factorisation. Taking both factors back out would double-count and give n2n^2 again, not nn. One factor per pair exactly undoes the doubling that squaring caused, which is the whole logic of the method.

What if a number has more than four digits — does prime factorisation still work?

Mathematically yes, but practically the method gets slow. Larger numbers tend to have bigger or more numerous prime factors, so the factor tree grows tall and easy to mess up by hand. That is exactly why the syllabus pairs this method with long division, which stays efficient for numbers of any size, whether or not they are perfect squares. Use prime factorisation when the number is small enough to factorise quickly and you also want to see its structure; switch to long division for anything larger.

How is finding the smallest multiplier different from finding the square root itself?

Finding x\sqrt{x} assumes xx is already a perfect square and asks you to pair up factors that are already complete. Finding the smallest multiplier starts from a number that is not a perfect square, spots which primes are left unpaired after factorising, and multiplies by exactly those leftover primes to complete every pair. Once you do that, the new, larger number is a perfect square, and you find its root the usual way — as in Worked example Step 5, where 14701470 needed a 3030 before its root could be taken.

Before this

Where this leads