In short
The sum of the first n terms of a geometric progression with first term a and common ratio r \neq 1 is S_n = a\,\dfrac{r^n - 1}{r - 1}. When |r| < 1 and you keep adding forever, the terms shrink to nothing and the infinite sum converges to S_\infty = \dfrac{a}{1 - r}. Both formulas come from the same trick: multiply the sum by r, shift everything by one position, and subtract to cancel almost every term.
A ball is dropped from a height of 18 metres. Each time it hits the ground, it bounces back to \dfrac{2}{3} of the height it fell from. So it rises to 12 metres after the first bounce, then 8 metres, then \dfrac{16}{3} metres, and so on. You want to know: what is the total distance the ball travels before it comes to rest?
The ball falls 18 metres, rises 12, falls 12, rises 8, falls 8, and so on. The upward distances alone form a GP: 12, 8, \dfrac{16}{3}, \dfrac{32}{9}, \dots with first term 12 and common ratio \dfrac{2}{3}. Each bounce is smaller than the last, and the total distance is finite — the GP converges. To find that total, you need a formula for summing the terms of a GP. That is what this article derives.
The finite sum formula
The trick: multiply and subtract
Take the GP a, ar, ar^2, \dots, ar^{n-1} and call its sum S_n:
Now multiply both sides by r:
Look at the two lines. The right-hand side of the second line is almost identical to the first — every term of S_n except the first appears in rS_n, and rS_n has one extra term (ar^n) at the end. Subtract the first equation from the second:
That is it. One multiplication, one subtraction, and all the intermediate terms cancel in pairs. The formula works for any r \neq 1. When r = 1, every term equals a and the sum is just na — no formula needed.
The two forms of the formula
Depending on whether r > 1 or r < 1, one form of the formula is more convenient than the other:
These are algebraically identical — multiply the top and bottom of either by -1 to get the other. Use whichever keeps the signs clean.
The infinite GP sum
What happens when |r| < 1
Take the finite sum formula and ask what happens as n \to \infty:
When |r| < 1, the term r^n shrinks as n grows. For example, \left(\tfrac{1}{2}\right)^{10} = \tfrac{1}{1024} \approx 0.001, and \left(\tfrac{1}{2}\right)^{20} \approx 0.000001. As n increases without bound, r^n \to 0, and the sum approaches
This is the sum to infinity of a GP. The infinite sum is finite because the terms shrink fast enough that their total cannot grow without bound.
When |r| \geq 1, the term r^n does not shrink. If r = 1, every term is a and the sum grows without bound. If r = -1, the sum oscillates (a, 0, a, 0, \dots) without settling. If |r| > 1, the terms themselves grow, and the sum diverges. The infinite GP sum exists only when |r| < 1.
Visualising the infinite sum with stacking
There is a beautiful geometric way to see why 1 + \dfrac{1}{2} + \dfrac{1}{4} + \dfrac{1}{8} + \cdots = 2. Start with a square of side length 1 (area = 1). Place a rectangle of area \dfrac{1}{2} next to it. Then a rectangle of area \dfrac{1}{4}, then \dfrac{1}{8}, and so on. Each piece is half the size of the previous one, and together they tile a rectangle of dimensions 2 \times 1 — total area 2.
Applications
Application 1: Recurring decimals
The recurring decimal 0.\overline{36} = 0.363636\dots is an infinite GP in disguise:
The first term is a = 0.36 and the common ratio is r = 0.01 (each block is shifted two decimal places to the right). Since |r| = 0.01 < 1, the infinite sum formula applies:
This recovers the fraction \dfrac{4}{11} from its decimal expansion — the same result the multiply-and-subtract trick from Number Systems gives, but through a different lens. Every recurring decimal is an infinite GP, and the sum-to-infinity formula is the machinery that converts it back to a fraction.
Application 2: Compound interest
You invest ₹P at an annual interest rate of r_{\%} compounded yearly. The balance at the end of year n is P(1 + r_{\%}/100)^n — a GP term. But suppose you also deposit ₹D at the end of every year. What is the total balance after n years?
Each deposit earns interest for a different number of years. The deposit made at the end of year 1 grows for n - 1 years to become D(1 + r_{\%}/100)^{n-1}. The deposit at the end of year 2 grows for n - 2 years. And the last deposit (year n) earns no interest.
The total from the deposits is
This is a finite GP with first term D, common ratio (1 + r_{\%}/100), and n terms. By the sum formula:
This is the formula banks use for Systematic Investment Plans (SIPs) and recurring deposits — a direct application of the GP sum.
Application 3: Bouncing ball
Return to the opening example. The ball is dropped from 18 metres and bounces to \dfrac{2}{3} of its previous height each time. The total distance is:
The factor of 2 accounts for both the rise and the fall of each bounce. The infinite GP in the brackets has a = 12 and r = \dfrac{2}{3}:
So the total distance is 18 + 2 \times 36 = 18 + 72 = 90 metres.
Two worked examples
Example 1: Find the sum of the first 8 terms of the GP $3, 6, 12, 24, \dots$
Step 1. Identify a and r.
Why: the first term is 3 and the common ratio is found by dividing consecutive terms.
Step 2. Choose the appropriate form of the formula. Since r = 2 > 1, use
Why: with r > 1, both the numerator and denominator come out positive, so the signs stay clean.
Step 3. Compute r^n.
Why: the exponent is n = 8, the number of terms. You can build this up: 2^4 = 16, so 2^8 = 16^2 = 256.
Step 4. Substitute into the formula.
Result: S_8 = 765.
The last term (384) is more than half the total (765). This is a characteristic feature of GPs with r > 1: the latest terms dominate the sum, because each term is larger than all the previous terms combined (in fact, for r = 2, each term equals the sum of all previous terms plus a).
Example 2: Find the sum to infinity of the GP $27, 9, 3, 1, \dfrac{1}{3}, \dots$
Step 1. Identify a and r.
Why: each term is one-third of the previous one.
Step 2. Check that |r| < 1.
Why: the infinite sum formula only applies when the terms are shrinking toward zero, which requires |r| < 1.
Step 3. Apply the infinite sum formula.
Why: dividing by a fraction is the same as multiplying by its reciprocal.
Step 4. Verify with partial sums. S_1 = 27. S_2 = 36. S_3 = 39. S_4 = 40. S_5 = 40.\overline{3}. The sums are indeed climbing toward 40.5 from below.
Result: S_\infty = \dfrac{81}{2} = 40.5.
The speed of convergence depends on |r|. When |r| is close to 0 (as here, r = \tfrac{1}{3}), the partial sums race toward the limit. When |r| is close to 1, convergence is painfully slow.
Common confusions
-
"The infinite sum formula works for any GP." Only when |r| < 1. If |r| \geq 1, the terms do not shrink to zero, the partial sums grow without bound (or oscillate), and the infinite sum does not exist.
-
"S_n = a\,\dfrac{r^n - 1}{r - 1} works when r = 1." It does not — substituting r = 1 gives \dfrac{0}{0}. When r = 1, every term is a, so the sum is just na.
-
"The sum of an infinite GP is the last term." There is no last term — the GP goes on forever. The sum is the limit of the partial sums, not a term of the sequence.
-
"If I keep adding positive terms forever, the total must be infinite." The bouncing ball and the tiling rectangle both show this is false. When each term is a fixed fraction of the one before, the total is finite. The key insight: a sum of infinitely many positive numbers can be finite, as long as the numbers shrink fast enough.
-
"The formula \dfrac{a}{1 - r} gives the exact sum after infinitely many terms." Yes, but "infinitely many terms" is a limit concept, not a finished process. No physical process ever completes infinitely many steps. The formula says that the partial sums get arbitrarily close to \dfrac{a}{1 - r}, and that no other number has this property.
Going deeper
If you came here to learn the finite and infinite GP sum formulas and see how they apply, you have them. The rest of this section is for readers who want to push into two directions: a deeper look at why the infinite sum converges, and a connection to a beautiful piece of mathematics.
Why does the infinite sum converge?
The formula tells you that S_\infty = \dfrac{a}{1 - r} when |r| < 1, but why is a sum of infinitely many positive numbers finite? The answer comes from looking at what is left over.
After n terms, the partial sum is S_n = a \cdot \dfrac{1 - r^n}{1 - r}. The "remainder" — everything you have not yet added — is
This remainder is a GP term scaled by \dfrac{1}{1 - r}. Since |r| < 1, the term r^n shrinks exponentially. After 10 terms of the GP with r = \tfrac{1}{2}, the remainder is \dfrac{a}{1024(1 - r)} — a thousandth of the full sum. After 20 terms, it is a millionth. The remainder vanishes not because you run out of terms, but because each remaining term is so small that even infinitely many of them cannot add up to anything significant.
This is the heart of convergence: the tail of the series is negligible, not because the terms are zero, but because they shrink so fast that their total can be made as small as you like.
Connection to \dfrac{1}{1 - x}
The finite GP sum can be written as
(take a = 1, r = x). When |x| < 1 and n \to \infty, the left side becomes the infinite series
This identity — the geometric series — is one of the most important formulas in mathematics. It expresses a simple rational function \dfrac{1}{1 - x} as an infinite sum of powers. The idea that a function can be written as an infinite polynomial (a power series) is the starting point of Taylor series, Fourier series, and much of analysis. Every one of those topics begins by generalising the geometric series to functions other than \dfrac{1}{1 - x}.
The GP sum formula is where this entire story starts.
The Koch snowflake's perimeter
Here is a famous application. Start with an equilateral triangle of side length 1. On each side, build a smaller equilateral triangle on the middle third. Repeat forever. The result is the Koch snowflake — a curve that encloses a finite area but has infinite perimeter.
At each step, the number of sides is multiplied by 4 and each side is divided by 3. The perimeter at step n is 3 \cdot \left(\dfrac{4}{3}\right)^n. Since \dfrac{4}{3} > 1, this GP diverges — the perimeter grows without bound. But the area added at each step forms a GP with ratio \dfrac{4}{9} < 1, so the total area converges. A shape with finite area and infinite boundary — geometry built from the convergence and divergence of GPs.
Where this leads next
The GP sum formula opens the door to infinite series, one of the central ideas of analysis and calculus.
- Geometric Progression — the foundation: the definition, the n-th term formula, and the properties of the sequence whose sum you have just computed.
- Geometric Mean — the multiplicative centre of a GP, and how it connects to the AM-GM inequality.
- Sum of Arithmetic Progression — the additive counterpart of this article, where the sum formula uses averaging rather than the multiply-and-subtract trick.
- Percentages and Ratios — compound growth, successive discounts, and multiplier chains all use the GP sum behind the scenes.
- Mathematical Induction — a rigorous way to prove the GP sum formula for all n, by showing it holds for n = 1 and that if it holds for n = k it holds for n = k + 1.