Significant Figures Explained: Counting the Digits That Matter

A lab balance reads 0.0010300 g. Which of those digits mean something, and which are padding? Significant figures are the counting system that answers this — the digits that carry real measurement information, as opposed to the zeros that merely hold the decimal point in place.

The four counting rules

Every sig-fig question reduces to four rules:

1. Non-zero digits are always significant        105    → 3
2. Zeros between non-zeros are significant       105    → 3
3. Leading zeros never are (they locate the
   decimal point)                                0.0010300 → 5
4. Trailing zeros count only after a decimal
   point                                         450 → 2   but  10.0 → 3

Rule 4 is where almost every mistake lives. The trailing zeros of 450 could be measured or could be placeholders, so the conservative reading counts only the 4 and 5. Write 450. with a decimal point — or better, 4.50 × 10² — and the zeros become significant declarations. Run the same number through a sig-fig calculator both ways and the counts of 2 versus 4 make the distinction visible.

Why the input is text, not a number

Here is a trap most calculators fall into: type 0.0010300 into an ordinary numeric field and it instantly becomes 0.00103 — the trailing zeros vanish before any counting starts, destroying exactly the information you wanted to analyze. A sig-fig tool therefore has to treat input as text. Type 0.0010300 and it counts 5 significant figures (1, 0, 3, 0, 0); type 1.20e3 and it reads the mantissa 1.20 for 3. The scientific-notation case is elegant: precision lives in the coefficient, so the exponent never disturbs the count.

Rounding to N significant figures

Counting is half the job; the other half is rounding a raw result down (or up) to the precision your data can support. Rounding starts from the leftmost significant digit and keeps N of them:

123456    → 3 s.f. → 123000        (1.23 × 10^5)
0.0010300 → 2 s.f. → 0.0010        (1.0 × 10^-3)
9999      → 2 s.f. → 10000         (1.0 × 10^4)

The last row is the famous surprise: rounding 9999 to 2 significant figures gives 10000, a result with more digits than the input. The value had to climb to the next power of ten, so the plain decimal form now shows four trailing zeros that are definitely not significant. This is precisely why the calculator pairs every rounded answer with its scientific-notation form — 1.0 × 10^4 states "two significant figures" unambiguously, while 10000 does not.

Sig figs versus scientific notation

The two ideas are easy to blur because they travel together, but they do different jobs. Scientific notation rewrites a number as coefficient × 10^exponent to manage magnitude and remove ambiguity about trailing zeros. Significant figures count and enforce how many digits carry precision, in whatever notation the number is written. You can have either without the other: 1200 is scientific-notation-ambiguous yet has a definite conservative count (2), while 1.20 × 10³ wears its count (3) on its sleeve. A scientific notation converter rewrites numbers; a sig-fig calculator audits them. For the full mechanics of the × 10^n format, see how scientific notation works.

The strange case of written zeros

What is the significant-figure count of 0 itself? There are no non-zero digits, so the rules go quiet. The working convention is to count decimal places: 0 counts as 0 significant figures, 0.0 as 1, and 0.00 as 2 — a recorded zero's precision is simply the number of places it was written to. It looks like a corner case, but it matters in labs, where "0.00" on a display is a real reading meaning "below this instrument's resolution," and the number of zeros is the honest description of what was measured.

Try it yourself

The Sig-Fig Calculator counts the significant figures in any number you type — trailing zeros and scientific notation included — and rounds it to a target count, showing both the plain and scientific-notation results. The Scientific Notation Converter rewrites numbers in unambiguous coefficient-exponent form, and the Exponent Calculator computes the powers of ten the notation leans on.