How Can π Have an Exact Value If Its Decimals Never End?

Here is the tension. Your textbook says \pi is the exact ratio of a circle's circumference to its diameter. Every circle in the universe has the same ratio, and that ratio is \pi. But then the textbook also says \pi = 3.14159265358979\dots with digits that never stop and never repeat. How can both be true? If you cannot write down all the digits, in what sense is \pi "exact"?

The resolution is that the decimal expansion and the number itself are two different things. The decimal is a representation. The number is a geometric fact about circles. The digits of \pi are infinite in the representation, but that is the decimal system's limitation — not \pi's.

What "exact" actually means here

A number is exact if it is a single, specific, fully-determined value. \pi is exact in this sense: there is one number — the ratio of any circle's circumference to its diameter — and that number is fixed, unambiguous, and the same for every circle.

\pi is not exact in the sense of "you can write down a finite decimal equal to it." But "writing down a finite decimal" is not what exactness means. Exactness is about whether the value is uniquely specified, not about how many digits it takes to write.

Compare: 1/3 is exactly specified. You know precisely what number it is. It also has an infinite decimal expansion: 0.333\dots The infinite decimal does not make 1/3 "approximate" — the decimal is just one way of writing it, and happens to be a way that does not terminate.

Three numbers, three kinds of decimal expansion, all three exact. The decimal system is good at representing some numbers finitely and others only as an infinite process. That is a feature of the decimal system, not a defect of the numbers.

How \pi is defined — no decimals involved

\pi is defined geometrically. Take any circle. Measure its circumference C and its diameter d. Compute C/d. You get the same number every time, regardless of the circle's size — the ratio does not care whether the circle fits on a coin or wraps around Jupiter. That universal ratio is \pi:

\pi = \frac{C}{d}.

Notice: this definition never mentions decimals. It says "measure a circle, take a ratio, that's \pi." The number is defined by circles. Decimals come later, when you want to write it down numerically.

Equivalently, \pi is the area of a unit circle — a circle of radius 1. That is also a purely geometric definition. It gives the same number, with no digit approximation involved.

So when someone says "\pi never ends," they are talking about a consequence of \pi being irrational — the decimal expansion of \pi does not terminate and does not repeat. They are not saying \pi itself is unfinished or incomplete. The circle has been drawn. The ratio has been taken. The number is there. Only its decimal representation is unending.

The decimal expansion is a description, not the number

Think of a number as a position on the number line. A decimal expansion is a procedure for zeroing in on that position — each additional digit halves or tenths the uncertainty.

For \pi, the procedure goes:

\pi is between 3 and 4. (First digit: 3.)
\pi is between 3.1 and 3.2. (Refined: 3.1.)
\pi is between 3.14 and 3.15. (Refined: 3.14.)
\pi is between 3.141 and 3.142. (Refined: 3.141.)
And so on, forever.

Each step zooms in by a factor of 10, pinning \pi's position tighter. But because \pi is irrational, no finite step ever lands exactly on it — the interval just keeps shrinking. The decimal is an infinite procedure. The number it describes sits at a single point on the line, and every step of the procedure gets you closer to that point.

Why the procedure never terminates: if \pi had a terminating decimal, say \pi = 3.14159265 \dots 0000\dots with only finitely many non-zero digits, then \pi would be a fraction with a power of 10 as the denominator, which would make \pi rational. The 1761 proof by Lambert showed \pi is irrational — no such fraction exists.

The same thing happens with \sqrt{2}, and no one objects

Nobody finds \sqrt{2} "approximate" just because its decimal 1.41421356\dots never ends. \sqrt{2} is defined geometrically too — it is the length of the diagonal of a unit square. The Pythagorean theorem gives it a specific, exact value. Its decimal expansion is irrational, so it never terminates and never repeats. And yet \sqrt{2} is an exact number.

\pi works the same way, just with circles instead of squares:

\sqrt{2} = diagonal of unit square. Exact geometric construction. Irrational decimal.
\pi = circumference over diameter of any circle. Exact geometric construction. Irrational decimal.

In both cases the number is exact. The decimal is infinite. Those are compatible statements.

What "approximation" really means

When you write \pi \approx 3.14 or \pi \approx 22/7, you are not saying \pi is fuzzy. You are saying: "the true value of \pi is not equal to 3.14, but it is close enough for some purposes." The \approx symbol admits the inexactness of your truncation, not of \pi itself.

\pi \approx 3.14 is correct to two decimal places. Error \approx 0.0016.
\pi \approx 3.1416 is correct to four decimal places. Error \approx 7.3 \times 10^{-6}.
\pi \approx 3.141592653589793 is the IEEE-754 double-precision value. Correct to about 15 decimal places.
NASA uses about 15 digits of \pi for interplanetary navigation — that is enough to place a spacecraft at Neptune's orbit with sub-centimetre accuracy. No calculation in science or engineering needs more than about 40 digits of \pi.
\pi itself has infinitely many digits. Trillions of them have been computed, mainly as a computing benchmark.

None of this changes the fact that \pi — the geometric ratio — is a single exact number. When you write \pi as a symbol in an equation, that symbol means that exact number. When you plug in 3.14 in a calculation, you are substituting an approximation because you cannot carry infinitely many digits through arithmetic. The distinction is important:

\text{area of circle of radius } r \;=\; \pi r^2 \qquad \text{(exact)}

\text{area of circle of radius } r \;\approx\; 3.14 \cdot r^2 \qquad \text{(approximate, with known error)}

In the first equation, the \pi is a symbol for the exact number. In the second, you have replaced the symbol with a truncated decimal, and the \approx admits that replacement.

Why $\pi = C/d$ is an exact relationship

Take a circle of diameter exactly 10 cm. What is its circumference?

C = \pi \cdot d = \pi \cdot 10 = 10\pi \text{ cm.}

That is the exact answer. The circumference is 10\pi cm — a specific real number. If someone asks "but what is it in decimals?" you can write:

C = 10\pi \approx 31.4159265\dots \text{ cm.}

The exact form is 10\pi. The decimal form is an approximation to as many digits as you want.

In JEE and board problems, leaving answers as \pi, \sqrt{2}, or e — rather than decimal approximations — is usually preferred, because those symbolic forms are exact. Converting to a decimal is a separate step that introduces rounding error.

The takeaway

\pi is exact because it is defined geometrically and uniquely specifies one point on the real line.
\pi's decimal expansion is infinite and non-repeating because \pi is irrational.
These two facts are independent. Many rationals have infinite decimals too (like 1/3), and they are still exact. Many irrationals have known exact forms (like \sqrt{2}, e, \pi), and they are still exact.
An "infinite decimal" is a statement about the decimal system's ability to represent a number in finitely many digits. It is not a statement about whether the number itself is well-defined.

A number exists independently of how you write it. The symbol \pi, the definition "C/d for any circle," and the digit string 3.14159\dots are three different ways of pointing at the same exact real number. The digit string is the least elegant of the three — but the number it is pointing at is as precise as mathematics can be.

This satellite sits inside Number Systems.