The binomial theorem expands an integer power of a two-term expression. It explains why the coefficients follow Pascal’s triangle and allows one selected coefficient to be found without writing the entire expansion.

The theorem is a finite identity for a nonnegative integer exponent. Infinite expansions for negative or fractional exponents belong to Power Series and require additional convergence conditions.

On this page

Derive the coefficients by counting choices

For a positive integer nn, the product (a+b)n(a+b)^n contains nn copies of (a+b)(a+b). Each term in the expanded product chooses either aa or bb from every copy.

To obtain an−kbka^{n-k}b^k, choose bb from exactly kk of the nn factors. There are (nk)\binom nk such selections. Collecting equal terms gives

(a+b)n=∑k=0n(nk)an−kbk,(a+b)^n=\sum_{k=0}^{n}\binom nk a^{n-k}b^k,

where (nk)=n!/[k!(n−k)!]\binom nk=n!/[k!(n-k)!] for integers 0≤k≤n0\le k\le n.

For n=0n=0, the polynomial expansion is the constant 1. Treat the formula as a polynomial identity: a factor with exponent zero contributes 1. This avoids interpreting a displayed endpoint term as an unrelated numerical question about 000^0.

The proof counts selections, not probabilities. It is valid for real or complex a,ba,b and does not require a+ba+b to be positive. The OpenStax binomial-theorem chapter gives the standard expansion and coefficient methods.

Use Pascal’s rule to organise the coefficients

The first rows of binomial coefficients are

1;1,1;1,2,1;1,3,3,1;1,4,6,4,1.1;\quad1,1;\quad1,2,1;\quad1,3,3,1;\quad1,4,6,4,1.

For 1≤k≤n−11\le k\le n-1, Pascal’s rule is

(nk)=(n−1k−1)+(n−1k).\binom nk=\binom{n-1}{k-1}+\binom{n-1}{k}.

To count kk-element selections from nn objects, separate those containing a chosen object from those not containing it. The two cases are disjoint and exhaust all selections, which proves the identity.

The edge coefficients are 1 because there is one way to select none or all of the objects. Symmetry (nk)=(nn−k)\binom nk=\binom n{n-k} follows by matching each selection with its complementary selection.

Pascal’s rule combines two disjoint selection cases

Adjacent selection counts combine in Pascal trianglePascal triangle shows rows zero through five. Two entries of three in row three feed the entry six in row four, expressing six equals three plus three.Binomial coefficients count selectionsn = 01n = 111n = 2121n = 31331n = 414641n = 515101051Choose the marked object, or do not: 3 + 3 = 6.
Rows are indexed from n=0. Edge entries equal one.
Diagram explanation and exact data

The six ways to choose two objects from four split into three choices containing one designated object and three not containing it. The arrows identify those two counts, not a probability flow.

Pascal’s rule combines two disjoint selection cases: exact reference data
Row n Coefficients
0 1
1 1, 1
2 1, 2, 1
3 1, 3, 3, 1
4 1, 4, 6, 4, 1
5 1, 5, 10, 10, 5, 1

Expand a signed binomial carefully

For (2x−3)4(2x-3)^4, use a=2xa=2x and b=−3b=-3, retaining the negative sign inside each power:

(2x−3)4=(2x)4+4(2x)3(−3)+6(2x)2(−3)2+4(2x)(−3)3+(−3)4.(2x-3)^4=(2x)^4+4(2x)^3(-3)+6(2x)^2(-3)^2+4(2x)(-3)^3+(-3)^4.

Simplification gives

16x4−96x3+216x2−216x+81.16x^4-96x^3+216x^2-216x+81.

The alternating signs come from the powers of −3-3. They are not supplied by changing the coefficients in Pascal’s triangle. A constant-term check gives (−3)4=81(-3)^4=81, and a leading-term check gives (2x)4=16x4(2x)^4=16x^4.

At x=1x=1, the coefficients sum to 16−96+216−216+81=116-96+216-216+81=1, matching (2−3)4(2-3)^4. These checks are useful, but the term-by-term derivation establishes the entire polynomial identity.

Find one coefficient without expanding everything

Find the coefficient of x5x^5 in (2x−3)8(2x-3)^8. Its general term, indexed by kk, is

Tk+1=(8k)(2x)8−k(−3)k.T_{k+1}=\binom8k(2x)^{8-k}(-3)^k.

The power of xx is 8−k8-k. Set 8−k=58-k=5 to obtain k=3k=3. Then the required coefficient is

(83)25(−3)3=56⋅32⋅(−27)=−48384.\binom83 2^5(-3)^3=56\cdot32\cdot(-27)=-48384.

This is the fourth term in the conventional descending-power order, because its index is k+1=4k+1=4. The value k=3k=3 counts selections of the second binomial term; it is not the ordinal position of the term.

Solve the exponent equation for a constant or missing term

Consider

(x2+1x)6,x≠0.\left(x^2+\frac1x\right)^6,\qquad x\ne0.

The general term is

(6k)(x2)6−kx−k=(6k)x12−3k.\binom6k(x^2)^{6-k}x^{-k}=\binom6k x^{12-3k}.

A constant term requires 12−3k=012-3k=0, so k=4k=4. Its value is (64)=15\binom64=15.

Now ask for an x5x^5 term in (x2+1/x)8(x^2+1/x)^8. The exponent equation is 16−3k=516-3k=5, which gives k=11/3k=11/3. That is not an integer index from 0 through 8. Therefore the coefficient of x5x^5 is zero: no such term occurs.

The presence of reciprocals makes these Laurent-polynomial expressions rather than ordinary polynomials in xx, but the binomial expansion remains a finite identity on x≠0x\ne0. Preserve that original restriction.

Use an exact expansion to assess an approximation

For 1.0251.02^5, write 1.02=1+u1.02=1+u with u=0.02u=0.02. The finite identity gives

(1+u)5=1+5u+10u2+10u3+5u4+u5.(1+u)^5=1+5u+10u^2+10u^3+5u^4+u^5.

Substituting u=0.02u=0.02 gives

1.025=1.1040808032.1.02^5=1.1040808032.

The first-order approximation 1+5u=1.11+5u=1.1 is below the exact result by 0.0040808032. Its direction is known because every omitted term is positive for this positive uu.

Keeping the quadratic term gives 1.104, with error 0.0000808032. Calling either expression an approximation rather than an equality preserves the distinction between truncating an expansion and using the full finite identity.

For negative uu, omitted terms can have different signs. Do not carry over a positive-term error argument unchanged. For a noninteger exponent, there is generally no terminating expansion, so an error bound requires a different argument.

Make the coefficient checks meaningful

For P(x)=(2x−3)4P(x)=(2x-3)^4, the sum of its coefficients is P(1)=1P(1)=1. The alternating sum, obtained by substituting x=−1x=-1, is P(−1)=625P(-1)=625. The coefficient magnitudes also sum to 625 here because the signs alternate in descending powers for this even exponent.

More generally, the sum of all binomial coefficients in row nn is 2n2^n, obtained by setting a=b=1a=b=1. For n≥1n\ge1, their alternating sum is zero, obtained by setting a=1a=1, b=−1b=-1.

These are evaluations of an identity with well-defined inputs. They are useful for detecting arithmetic errors and for connecting coefficient patterns to counting subsets, as developed in Combinatorics.

Practice questions

  1. Expand (x+2)3(x+2)^3.
  2. Find the coefficient of x5x^5 in (2x−3)8(2x-3)^8.
  3. Find the constant term in (x2+1/x)6(x^2+1/x)^6, and state its domain restriction.
Check your answers
  1. The coefficients 1,3,3,11,3,3,1 give x3+3x2(2)+3x(4)+8=x3+6x2+12x+8x^3+3x^2(2)+3x(4)+8=x^3+6x^2+12x+8.
  2. Set 8−k=58-k=5, giving k=3k=3. The coefficient is (83)25(−3)3=−48384\binom83 2^5(-3)^3=-48384.
  3. The exponent is 12−3k12-3k, so k=4k=4. The constant term is (64)=15\binom64=15. The original expression requires x≠0x\ne0 even though this one term is constant.

Related topics

Combinations explains the selection counts in the coefficients. Power Series and Taylor and Maclaurin Series distinguish finite binomial identities from infinite local representations.

Sources and further reading