Which Zeros Are Significant? Every Case, Settled.

A zero is significant when it does real work reporting precision: sitting between two non-zero digits, or trailing after a decimal point where someone confirmed that position is exactly zero. A zero is not significant when it only marks where the decimal point falls (a leading zero). And a trailing zero in a plain whole number with no decimal point — like the zeros in 4500 — is genuinely ambiguous on its own; that ambiguity is a real limitation of decimal notation, not a trick question, and it’s exactly what scientific notation was built to fix.

Every non-zero digit in a number is automatically significant — nobody argues about whether the “7” in 470 counts. Zero is the only digit that ever needs a rule to settle its status, which is why it causes almost all of the confusion in this topic. Our Complete Guide to Rules, Zeros, and Exceptions covers the five general rules at a summary level; this page goes through every zero scenario in full, including a couple of edge cases — the “100.” convention, and what happens to zeros after a unit conversion — that most references skip entirely.

DEV NOTE: Embed live calculator widget here. Shortcode: [sfc_calculator mode=”mini” default_input=”1200″ show_rule_panel=”true” highlight_zeros=”true”]. The highlight_zeros flag should color-code every zero in the input by case (leading/captive/trailing-significant/trailing-ambiguous) using the same palette as the B12 digit-map graphic below, so the live tool and the static content use identical visual language.

[Live calculator embeds here] — Enter any number to see every zero in it color-coded by case, with the rule that applies to each one.

The 7 Zero Cases

Case 1 — Leading zeros: never significant. Any zero before the first non-zero digit exists only to mark where the decimal point falls. It carries no information about how precisely the number was measured. (Example 1)

Case 2 — Captive zeros: always significant, no exceptions. A zero sitting between two non-zero digits always counts — including strings of several captive zeros in a row. There is no version of this rule with a caveat. (Example 2)

Case 3 — Trailing zeros with a decimal point present: always significant. This holds regardless of the number’s magnitude — whether the whole part is 45 or 0. A zero written after the decimal point, following the last non-zero digit, is there because someone confirmed that position is exactly zero, not because it’s a placeholder. (Example 3)

Case 4 — Trailing zeros with no decimal point: genuinely ambiguous. This is the one real weak spot in plain decimal notation. A whole number’s trailing zeros could reflect an exact measurement or could just be unknown placeholders down to that precision, and the written digits alone don’t say which. NIST’s own SI style guide uses precisely this example: written as 1200 m, it’s simply not possible to tell from the digits whether the last two zeros are significant or only indicate magnitude. (Example 4)

Case 5 — Trailing zeros inside scientific notation: always significant. This is exactly why scientific notation exists — writing a number in the form a.bcd × 10ⁿ makes every digit in the mantissa a deliberate, precision-bearing choice. There’s no ambiguity left to resolve. (Example 5)

Case 6 — The standalone zero: one significant figure, by convention. This case comes up rarely and matters little in practice — a bare “0” is nearly meaningless without a stated uncertainty. More useful to remember: a zero that results from a calculation should still carry whatever decimal places the calculation’s own rules require. (Example 6)

Case 7 — Zeros created by unit conversion: not a simple digit-matching rule. Converting a measurement using an exact factor (say, feet to meters) does not mean “keep the same number of significant figures as before” in every case. What actually has to be preserved is the original measurement’s relative precision — and matching that can mean keeping a different number of digits than a naive sig-fig count would suggest. (Example 7)

The one-line decision path, if you just need the shortcut: Is the zero before the first non-zero digit? Not significant. Is it between two non-zero digits? Significant. Is it after the last non-zero digit? Significant if a decimal point is present anywhere in the number (or you’re already in scientific notation) — ambiguous if not.

Worked Examples — One Per Case

Example 1 — Leading zeros (Case 1)

Number: 0.007060

The “0.00” only marks the decimal position. From the first non-zero digit onward: 7, 0, 6, 0 — the trailing zero counts because a decimal point is present.

Answer: 4 significant figures.

Example 2 — Captive zeros (Case 2)

Number: 40008

Every digit sits between the leading 4 and the trailing 8, so nothing here is a leading or trailing zero — all three zeros are captive and all count: 4, 0, 0, 0, 8.

Answer: 5 significant figures.

Example 3 — Trailing zeros with a decimal point, at two magnitudes (Case 3)

Numbers: 3.20 and 0.0320

3.20 → the decimal point is present, so the trailing zero after the “2” counts: 3, 2, 0. 0.0320 → the leading “0.0” isn’t counted, but from the first non-zero digit onward the same logic applies: 3, 2, 0.

Answer: 3 significant figures, both times. The rule doesn’t change with the number’s size — only the presence of the decimal point and the position relative to the first non-zero digit matter.

Example 4 — Ambiguous trailing zeros (Case 4)

Number: 45000, written with no decimal point and no other context

The two non-zero digits (4, 5) must count. Each of the three trailing zeros might or might not be significant, depending on what was actually measured:

Answer: anywhere from 2 to 5 significant figures — genuinely undecidable from “45000” alone. Whoever recorded this number needs to specify which one they meant.

Example 5 — Trailing zeros in scientific notation (Case 5)

Number: 3.00 × 10⁵

Every digit in the mantissa — 3, 0, 0 — was placed there deliberately.

Answer: 3 significant figures, unambiguously.

Example 6 — The standalone zero (Case 6)

Calculation: 5.00 g − 5.00 g

The raw result is 0, but the addition/subtraction rule still applies: match the fewest decimal places among the inputs, which is 2.

Answer: 0.00 g — not bare “0.” The trailing zeros here communicate how precisely you know the result is zero; they aren’t optional decoration.

Example 7 — Zeros created by unit conversion (Case 7)

Calculation: Convert 36 ft to meters, using the exact factor 1 ft = 0.3048 m

Raw product: 36 × 0.3048 = 10.9728. The naive approach — “36 has 2 sig figs, so keep 2” — gives 11 m. But NIST’s own worked example in its SI conversion guide shows this loses information: the relative rounding error of “36” is about ±1.4%, while “11” carries a relative error of about ±4.5% — more than three times worse. Rounding instead to 11.0 m (3 sig figs) keeps the relative error at about ±0.45%, close to the original measurement’s precision without overstating it.

Answer: 11.0 m — three significant figures, not two, because what matters after a conversion is preserving the original’s relative precision, not mechanically matching its digit count.

Where Even the Rules Get Tested

• One extra zero changes everything. 0.07 has 1 significant figure; 0.070 has 2. It’s a single keystroke, and it’s the difference between “roughly a tenth” and “measured to the nearest thousandth.”
• The decimal-comma trap. Many countries write 1,200 to mean one thousand two hundred, and others write 1.200 to mean the same thing — while yet others use exactly those punctuation marks with the opposite meaning (1.200 as one-point-two-zero-zero, with the period as the decimal marker). If you’re reading data from an international source, confirm which convention it uses before you count a single zero.
• Overline and underline notation don’t survive copy-paste. Some texts mark an ambiguous trailing zero as significant with a bar over it, or underline the last significant digit. Both work fine on a printed page and both routinely vanish when the number is copied into an email, a spreadsheet cell, or a text message — silently reintroducing the exact ambiguity the notation was meant to remove. Scientific notation doesn’t have this failure mode, which is the real reason it has mostly replaced overlines in modern usage.
• Not every zero-containing string is a measurement. A product code, a phone number, or an ID number can be full of zeros with no significant-figures meaning at all. Sig fig rules apply to numbers representing measured or calculated quantities — check that you’re looking at one before you start counting.
• Different tools resolve Case 4 differently by default — see the comparison below, because this one catches people off guard constantly.

How Different Tools and Contexts Handle the Ambiguous Case

Case 4 — a whole number’s trailing zeros with no decimal point — is where “how many sig figs does this have” stops being a question with one universal answer. Here’s how it typically gets resolved in practice. These are general tendencies, not universal rules — when it matters, check the specific source or tool’s own documentation.

This is also exactly why a calculator’s answer and a textbook’s answer can legitimately disagree on a problem involving a number like 100 — neither is “wrong,” they’re using different default assumptions about a case that plain notation leaves genuinely open.

Where the Rules Come From

The ambiguous-trailing-zero problem isn’t a classroom invention — it shows up in NIST’s own official style guide for using the SI, which uses almost the identical example given in Case 4 above to illustrate why plain notation can’t settle the question on its own. The unit-conversion nuance in Case 7 comes from the same document’s worked guidance on rounding converted values. Neither is a rounding-standard deep dive in the way ASTM E29 and the GUM are — those govern rounding procedure once a value’s precision is already known. This page is about the narrower, upstream question: figuring out what a number’s zeros are actually telling you before any rounding starts.

Common Mistakes

1. Assuming all trailing zeros in a whole number are automatically insignificant. This is the mirror image of assuming they’re all significant — both blanket assumptions are wrong. Case 4 numbers are genuinely ambiguous, not secretly one-sided.
2. Dropping a confirmed trailing zero when transcribing data — turning a measured 3.20 into 3.2 and silently discarding real precision information in the process.
3. Treating “0” as having no significant figures, or as undefined. By convention it has one — see Case 6.
4. Missing a regional decimal-comma swap when working with international data or literature.
5. Treating the zeros created by an exact unit conversion as automatically matching the original sig fig count, when the correct answer depends on relative error — see Case 7 and Example 7.
6. Losing overline or underline notation when copying a number between documents, formats, or devices.

Practice Problems

Concept: Leading zeros

Q1. How many significant figures are in 0.0080? A) 1 B) 2 C) 3 D) 4 Answer: B) 2. The leading zeros aren’t counted; 8 and the trailing 0 (decimal point present) both count.

Q2. How many significant figures are in 0.36? A) 1 B) 2 C) 3 D) 4 Answer: B) 2. The “0” before the decimal point is a leading zero and isn’t counted; only 3 and 6 are significant.

Concept: Captive zeros

Q3. How many significant figures are in 40008? A) 2 B) 3 C) 4 D) 5 Answer: D) 5. All three zeros sit between non-zero digits, so all five digits count.

Q4. How many significant figures are in 2.005? A) 1 B) 2 C) 3 D) 4 Answer: D) 4. Both zeros are captive, between the 2 and the 5.

Concept: Trailing zeros with a decimal point

Q5. How many significant figures are in 3.200? A) 2 B) 3 C) 4 D) 5 Answer: C) 4. The decimal point is present, so all three trailing digits after the “3” count.

Q6. How many significant figures are in 0.0500? A) 1 B) 2 C) 3 D) 4 Answer: C) 3. The leading zeros aren’t counted; 5 and the two trailing zeros are (decimal point present).

Concept: Ambiguous trailing zeros

Q7. Written as 90000 with no other context, what’s the most defensible statement about its significant figures? A) Definitely 1 B) Definitely 5 C) Ambiguous — could be 1 to 5 D) Definitely 2 Answer: C. Without a decimal point or scientific notation, this is genuinely undecidable from the digits alone.

Q8. Written as 3000 with no other context, what are the minimum and maximum plausible significant figure counts? A) Min 1, max 4 B) Min 1, max 3 C) Min 2, max 4 D) Min 4, max 4 (fixed) Answer: A. At minimum, only the “3” is confirmed significant; at maximum, all four digits could be exact.

Concept: Zeros from unit conversion

Q9. A rough measurement of 5 kg (1 significant figure) is converted using the exact factor 1 kg = 1000 g, giving a raw value of 5000 g. How should this be reported? A) 5000 g (4 sig figs) B) 5 × 10³ g (1 sig fig) C) 500 × 10¹ g D) 5.000 × 10³ g Answer: B. The conversion factor is exact and doesn’t add precision; the result should still reflect the original measurement’s 1 significant figure.

Q10. In the 36 ft → meters conversion from Example 7, why is the answer reported as 11.0 m (3 sig figs) rather than 11 m (2 sig figs, naively matching “36”)? A) Because 11 m rounds incorrectly B) Because 3 sig figs happens to look better C) Because 11 m’s relative error is much larger than the original measurement’s D) Because meters always get 3 sig figs Answer: C. Matching relative error, not digit count, is what the conversion actually requires — 11 m would discard real precision that 36 ft carried.

Zero Cases at a Glance

DEV NOTE: Render as an annotated-number graphic — green highlight for significant, grey for not significant, amber/striped for the genuinely ambiguous case. Text version below is the content spec for the design/dev team, not final reader-facing copy.

0.00[7][0][6][0] → bracketed digits significant → 4 significant figures (Case 1 + Case 3 combined)

[4][0][0][0][8] → every digit bracketed, none are leading or trailing → 5 significant figures (Case 2)

45000 → four legitimate readings depending on intended precision:

• 4.5 × 10⁴ (amber/ambiguous as written) → 2 sig figs
• 4.50 × 10⁴ → 3 sig figs
• 4.500 × 10⁴ → 4 sig figs
• 4.5000 × 10⁴ → 5 sig figs

Quick Reference

Continue Learning

DEV NOTE: Bake into the WikiWriter import payload at publish time per the music-dictionary.org rule.

Back to the fundamentals:

• Significant Figures: The Complete Guide to Rules, Zeros, and Exceptions

Go deeper on one case at a time:

• Are Leading Zeros Significant? (No — Here’s Why)
• Are Trailing Zeros Significant? It Depends on the Decimal Point
• Captive Zeros: Why Zeros Between Digits Always Count
• Why 1200 Can Have 2, 3, or 4 Significant Figures
• The Overline (Bar) Notation for Ambiguous Trailing Zeros
• Exact Numbers and Why They Never Limit Precision

Related topics:

• Scientific Notation: Complete Guide
• Rounding Standards Compared: ASTM E29, ISO 80000-1, NIST

Tools:

• Significant Figures Calculator
• Sig Fig Counter — with zero-highlighting built in

Sources and Further Reading

• NIST Guide to the SI, Chapter 7, Rules and Style Conventions for Expressing Values of Quantities — NIST’s own official style guide, which directly addresses the ambiguous-trailing-zero problem using an equivalent example to Case 4 above. (nist.gov)
• NIST Guide to the SI, Appendix B, Conversion Factors — the source of the Case 7 / Example 7 reasoning on rounding converted values by relative error rather than simple digit-matching. (nist.gov)
• A2LA, Figuring Out Significance: What Are Significant Figures — a laboratory accreditation body’s explainer connecting these definitions to NIST SP 811 §7.9 and everyday lab practice. (a2la.org)

Review and Methodology

DEV NOTE: Do not publish with this placeholder live — recruit a named, credentialed reviewer before this page goes public, per the modified PumpCalcs.com-rule gate already applied to Pillar 1.

Reviewed by: [Pending — reviewer assignment required before publication] Last reviewed: [Pending] Methodology: Every case above is cross-checked against NIST’s own SI style guide (see Sources). Calculator results referenced on this page use an arbitrary-precision decimal engine, not native floating-point math, validated against the site’s versioned regression fixture set.

Changelog

v1.0 — Initial draft completed, 2026-08-10.