GMT vs UTC

The confusion between GMT and UTC is among the most common time-related questions. They're close enough that most people use them interchangeably, but distinct enough that scientific, navigational, and legal contexts require specificity. This article covers the historical origin of each, the technical distinction, the legal status, and when the difference actually matters.

This is a companion to /learn/utc-explained and /learn/time-zones-explained.

GMT: the historical origin

Greenwich Mean Time (GMT) has its roots in the founding of the Royal Observatory Greenwich by King Charles II in 1675. The observatory's purpose was to support marine navigation by producing accurate astronomical tables for the longitude problem — see /learn/the-longitude-problem for context.

GMT is mean solar time at the Greenwich meridian. Mean solar time averages out the seasonal variation in the sun's apparent motion (the equation of time, which varies by up to ±16 minutes through the year). Mean solar time is the basis of civil clocks.

The Royal Observatory used a transit telescope to observe the moment when the sun crossed the local meridian — this was apparent solar noon. Averaged over the year, the sequence of apparent noons defined mean solar noon. GMT's seconds were defined by dividing the mean solar day into 24 × 60 × 60 = 86,400 equal parts.

GMT was adopted as the British legal time reference by the Definition of Time Act in 1880. The 1884 International Meridian Conference adopted the Greenwich meridian as the international prime meridian, which made GMT (or its astronomical equivalent) the de facto world reference for marine navigation and timekeeping.

UTC: the metrological replacement

By the mid-20th century, atomic clocks (cesium beam clocks, first operating in 1955 at the UK National Physical Laboratory) offered timekeeping more stable than the Earth's rotation. The SI second was redefined in 1967 in terms of cesium-133 hyperfine transitions (9,192,631,770 Hz), uncoupling the second from astronomical observation.

Coordinated Universal Time (UTC) was formally introduced on January 1, 1972, as the technical replacement for GMT in international science and telecommunications. UTC is defined by ITU-R Recommendation TF.460-6 and coordinated by the BIPM (Bureau International des Poids et Mesures).

UTC has two components:

1. TAI (International Atomic Time) — a continuous atomic-clock-based timescale computed from ~400 atomic clocks at ~80 laboratories worldwide.
2. Leap seconds — occasional one-second adjustments added to UTC (not TAI) to keep UTC within 0.9 seconds of UT1 (the astronomical timescale based on Earth's actual rotation).

See /learn/utc-explained for the detailed mechanism.

The technical difference

The fundamental difference: what the time is based on.

| Feature | GMT | UTC | | ------------------- | -------------------------------- | ------------------------------------ | | Basis | Earth rotation (astronomical) | Atomic clocks + leap seconds | | Origin (formal) | 1675 (Royal Observatory founded) | 1972 (ITU-R TF.460) | | Defining body | Royal Observatory Greenwich | BIPM, ITU | | Precision | ~1 second (Earth-rotation noise)| ~1 nanosecond (atomic-clock stability)| | Leap seconds | No (Earth rotation has none) | Yes (occasionally, to track UT1) | | Current source | UT1, derived from VLBI | TAI minus leap seconds | | Legal status (UK) | Reference for UK civil time | Reference for technical and scientific |

In modern practice, “GMT” in civil contexts typically means UT1 — the astronomical time scale — but is sometimes loosely used to mean UTC. The Royal Observatory itself publishes UTC under the “GMT” label for public consumption (BBC pips, broadcast time signals, etc.), which propagates the loose equivalence.

UT1 and UTC differ by at most 0.9 seconds at any moment. For most practical purposes, this is undetectable.

UT0, UT1, UT2: the UT family

To complete the picture, the “Universal Time” family of astronomical timescales:

• UT0: the raw mean solar time computed at a specific observation site. UT0 includes the effects of polar motion (the wobble of Earth's rotation axis relative to the crust).
• UT1: UT0 corrected for polar motion. UT1 is the global astronomical mean solar time at Greenwich; it reflects Earth's actual rotation. UT1 is what GMT means in modern technical usage.
• UT2: UT1 corrected for seasonal variation in Earth's rotation rate. UT2 is now obsolete; UT1 is the standard astronomical timescale.

The IERS in Paris computes and publishes UT1 via VLBI observations of distant quasars. The accuracy of modern UT1 measurements is sub-millisecond.

Legal status

UTC has no automatic legal status in any single country; it's an international scientific reference. Each country has its own legal time:

• United Kingdom: GMT (or British Summer Time when DST is in effect) is the legal time, per the Interpretation Act 1978 (Schedule 1).
• United States: each U.S. state designates its own legal time; many reference “UTC” or “Coordinated Universal Time” directly in modern law.
• France: legal time is “the time relevant to the meridian of Paris,” defined since 1911 as Paris Mean Time minus 9 minutes 21 seconds — which is GMT (and effectively UTC).
• Most other countries: reference UTC or local time derived from UTC, sometimes with country-specific legal-time legislation.

The UK's continued reference to GMT in legal documents is a historical artifact. UK courts and statutory deadlines reference GMT, even though the actual time signal distributed (and the time on every UK clock) is UTC for all practical purposes. The legal-versus- technical distinction has no day-to-day implications.

When the distinction matters

For most uses, GMT and UTC are interchangeable. The distinction matters in three contexts:

1. Sub-second precision applications

Scientific instruments, GPS receivers, telecom signaling, financial trading timestamps, and similar applications require timestamps accurate to milliseconds or better. At this precision:

• The 0.9-second potential offset between GMT (= UT1) and UTC matters.
• GPS receivers display UTC (computed from broadcast GPS Time + leap-second offset), not GMT.
• Network Time Protocol (NTP) distributes UTC, not GMT.
• IEEE 1588 Precision Time Protocol (PTP) distributes UTC for nanosecond-precision applications.

2. Astronomy

Astronomical observations are timestamped to milliseconds or better. A telescope observation logged as “GMT 2026-05-24 17:30:00.000” is ambiguous — does this mean UTC, UT1, or a loose civil interpretation? The International Astronomical Union (IAU) recommends explicitly stating “UTC” or “UT1” in published timestamps.

Historical astronomical data is particularly tricky: pre-1972 observations were timestamped in GMT (= UT1 implicit); modern observations are in UTC. Comparing across the boundary requires the leap-second history.

3. UK legal interpretation

UK statutory deadlines, court orders, and parliamentary proceedings reference GMT. In rare cases — e.g., a court hearing scheduled at “17:00 GMT” — the legal reference is GMT, not UTC. In practice these are indistinguishable for human attendance, but the legal phrasing matters if challenged.

GMT in modern internet protocols

Several long-lived internet standards retain the literal string “GMT” for what is now technically UTC. This is conventional rather than precise — the timestamp value is UTC; the textual label is “GMT” for historical compatibility.

HTTP Date headers (RFC 7231 / RFC 9110): HTTP defines the Date header value in the “preferred” IMF-fixdate format with the literal “GMT” suffix:

Date: Sun, 24 May 2026 17:30:00 GMT

The RFC explicitly notes that the literal string “GMT” refers to UTC. Every HTTP-compliant client and server emits and parses this format with the “GMT” literal. Modern web frameworks (Node.js, Go net/http, Python aiohttp, etc.) all produce this literal even though they internally use UTC.

RFC 5322 (Internet Message Format): email message headers accept “GMT”, “UT”, “UTC”, and numerical offsets like “+0000”. The first three all denote the same zero-offset reference; modern email senders typically use “+0000”.

Cookie expiration: Set-Cookie headers use the same IMF-fixdate format with “GMT”:

Set-Cookie: id=abc; Expires=Sun, 24 May 2026 17:30:00 GMT

SQL standards: SQL:2003 and later allow either UTC or GMT in time-zone string literals; most database engines treat them as equivalent.

The persistence of “GMT” in protocol text is purely historical: when these standards were drafted (HTTP/1.0 in 1996, RFC 822/5322 lineage going back to 1982, etc.), “GMT” was the natural label for zero-offset time. Changing the literal would break billions of existing implementations for no functional benefit; the literal stays, and the meaning is now “UTC” in practice.

What about “GMT+0” or “UTC+0”?

In time-zone notation, “GMT+0”, “UTC+0”, and “UTC” all refer to the same civil time zone (Western European Time without DST). When you see “GMT+5” or “UTC+5” in an application, both mean “UTC plus 5 hours” — they're interchangeable as time-zone notation.

The notation predates the GMT-to-UTC transition; older applications and APIs use “GMT”, newer ones prefer “UTC”. Either is correct in time-zone-offset notation; the difference is cosmetic.

Common misconceptions

“GMT is the time in London right now.” Not always. London uses British Summer Time (BST) (UTC+1) from late March to late October. During those months, London is on BST = GMT+1 (equivalent to UTC+1), not GMT. GMT is London's winter time; BST is London's summer time.

“GMT and UTC always agree exactly.” They agree to within 0.9 seconds, but the actual difference fluctuates with Earth's rotation. For any timestamp precision worse than a second, they're identical; for sub-second-precision purposes, the distinction matters.

“UTC stands for Coordinated Universal Time, but why isn't the abbreviation CUT?” It's a diplomatic compromise. English speakers wanted CUT (Coordinated Universal Time); French speakers wanted TUC (Temps Universel Coordonné). UTC was chosen as neither-language-specific, sometimes glossed as “UTC (Coordinated Universal Time, Temps Universel Coordonné).”

“Atomic clocks measure GMT.” They don't. Atomic clocks measure TAI (or contribute to TAI), from which UTC is derived. The astronomical observations that define UT1 (≈GMT) are separate from the atomic-clock measurements that define TAI/UTC.

“The BBC pips are GMT.” The pips broadcast by the BBC reference UTC (with the BBC source-time signal calibrated against UTC, not GMT directly). The pips are labelled as GMT by convention but are technically UTC.

“GMT was abolished in 1972.” It wasn't. UTC replaced GMT as the technical/scientific reference in 1972, but GMT continues as a legal and popular term in the UK and elsewhere. The Royal Observatory still maintains the meridian and the cultural association with timekeeping.

“GMT is more accurate than UTC.” UTC is vastly more precise (nanosecond accuracy from atomic clocks) than GMT/UT1 (millisecond from VLBI; second from classical astronomical observation). UTC's precision is one reason it replaced GMT for technical purposes.

“The 0.9-second tolerance between UTC and UT1 is arbitrary.” It was chosen as a compromise between metrologists (who wanted larger tolerances to reduce leap- second frequency) and navigators (who wanted tighter tolerances to keep UTC close to astronomical time). The 2022 CGPM Resolution 4 will increase the tolerance by or before 2035, effectively suspending leap seconds.