What Is a Good Latency? Ping Values by Use and Distance

Quick answer: there is no single "good" latency. A good result is low for the use and low for the distance. As a rough guide: under 30 ms is excellent for any activity with a nearby server, up to about 60 ms is good for competitive gaming, up to about 100 ms is fine for browsing, streaming and video calls, and above about 150 ms to a nearby server is a sign that something is wrong. At the same time, a server on the other side of the planet cannot respond in under roughly 150 ms however good your connection is, so for far destinations a result of 200 to 300 ms may be normal.

Important: ping is not the same number as GlobeLatency's HTTPS latency. The ranges on this page describe round-trip time to the server you actually use (what ping, or a game's own ping display, shows). GlobeLatency times full HTTPS requests from the browser, which include DNS, TCP and TLS setup on a fresh connection (roughly three round trips) plus server and browser time, so its numbers are usually higher than ping to the same server, often 1.5 to 3 times on a cold connection. Judge GlobeLatency results by the colour and the comparison with the world average, and use these ranges for ping or in-game latency. See ping vs HTTPS latency.

What is a good ping for each use?

Latency here means round-trip time (RTT) to the server you are actually using. See what is latency if the term is new. The figures below are practical guidelines, not standards; they differ by application, implementation and personal sensitivity.

Use case	Excellent	Acceptable	Problematic
Web browsing	Under 50 ms	50 to 150 ms	Over about 300 ms (pages start noticeably slowly)
Email, messaging, documents	Under 100 ms	100 to 300 ms	Over 500 ms
Video calls	Under 100 ms	100 to 250 ms	Over about 300 ms RTT, or unstable
Voice calls (VoIP)	Under 100 ms	100 to 250 ms	Over about 300 ms RTT
Video streaming	Under 150 ms	Up to a few hundred ms	Mostly a throughput problem; buffering hides delay
Cloud gaming	Under 30 ms	30 to 60 ms	Over about 80 ms feels delayed
Remote desktop, SSH	Under 50 ms	50 to 150 ms	Over 200 ms feels sluggish while typing
Financial trading (retail)	Under 50 ms	50 to 150 ms	Depends on strategy
High-frequency trading	Microseconds	Sub-millisecond	Any milliseconds

Gaming by genre

Genre	Comfortable	Noticeable	Hard to play
Competitive first-person shooters	Under 40 ms	40 to 80 ms	Over about 100 ms
Fighting games (rollback netcode)	Under 50 ms	50 to 100 ms	Over about 120 ms
MOBA and battle royale	Under 60 ms	60 to 100 ms	Over about 150 ms
MMO and RPG	Under 100 ms	100 to 200 ms	Over about 300 ms
Turn-based and strategy	Under 200 ms	200 to 500 ms	Rarely a barrier

Games also depend on jitter and packet loss. A steady 70 ms often plays better than an average of 40 ms that swings between 15 and 150. See latency for gaming for more.

What is a good latency by distance?

Distance sets a floor you cannot go beneath. The physical minimum assumes a perfectly straight fibre path and nothing else; real paths are 20 to 60 percent longer or more, and equipment adds more. The typical range is what you would reasonably see on a healthy connection.

Distance to server	Example	Physical minimum RTT	Typical healthy RTT
Same city (under 50 km)	Home to a local data centre	Under 1 ms	5 to 20 ms
Same region (about 300 km)	Neighbouring large cities	About 3 ms	10 to 30 ms
Same country (about 1,000 km)	Two cities in a mid-size country	About 10 ms	15 to 40 ms
Large country or continent (about 4,000 km)	US coast to coast	About 40 ms	55 to 90 ms
Across an ocean (about 6,000 km)	Western Europe to the US east coast	About 60 ms	70 to 100 ms
Far continent (about 10,000 km)	Europe to east Asia	About 100 ms	150 to 250 ms
Near antipodal (about 15,000 to 20,000 km)	Western Europe to New Zealand	150 to 200 ms	250 to 330 ms

The rule behind the table: roughly 1 ms of RTT per 100 km of fibre path, because light in fibre covers about 200,000 km/s. See undersea cables and the speed of light for why routes bend and which ones are short.

What is the minimum physically possible latency?

For a given straight-line distance d in kilometres:

• In a vacuum or air at the speed of light (about 300,000 km/s): RTT ≥ 2 × d ÷ 300,000 s, or about 0.67 ms per 100 km.
• In optical fibre (about 200,000 km/s): RTT ≥ 2 × d ÷ 200,000 s, or exactly 1 ms per 100 km.

Half the circumference of the Earth is about 20,000 km, so even a perfect fibre straight through the planet (not possible) would give about 200 ms RTT, and a surface-following cable more. Using the speed of light in vacuum it would be about 133 ms, which is the bound that radio links, low-orbit satellite constellations and some experimental hollow-core fibres try to approach. Geostationary satellites orbit at about 35,786 km, so a request that goes up and down and then the reply goes up and down covers about four times that, at least about 480 ms, before any processing.

Why is "good" relative to the destination?

Take a result of 120 ms.

• To a server in your own city, it is a clear problem: expect 5 to 20 ms.
• To a server on another continent, it is a good result.
• To an anycast network, which answers from a nearby edge, 120 ms is odd because those usually answer in a few tens of milliseconds.

Judging a number without the destination is therefore guesswork. The useful questions are: how far is it? (what is the physical minimum), how do other people see it? (what is typical for that server), and how consistent is it? (stable or varying).

How GlobeLatency judges your numbers

GlobeLatency compares your average for each destination with the world average of the same destination, collected from many visitors' tests. It divides your average by the world average and labels the result: at or below 0.7 of the average is great, up to 1.15 is good, up to 1.5 is slightly slower, and above 1.5 is clearly slower. Two exceptions avoid false alarms: a difference of under 30 ms, or a value under 50 ms, is never called bad. Only when no world data exists for a destination does it use fixed bands (under 50 ms great, under 120 ms good, under 250 ms mid, otherwise bad). Details are in global comparison explained and reading your results. You can try it with the free latency test.

Worked examples

Example 1: Should 85 ms to a game server worry me?

If the server is about 1,000 km away, a healthy RTT is 15 to 40 ms, so 85 ms is two to three times higher than expected and worth investigating (Wi-Fi, a congested line, a poor route). If the server is 6,000 km away across an ocean, 85 ms is perfectly normal.

Example 2: Comparing two destinations

You measure 28 ms to a destination whose world average is 30 ms, and 190 ms to another whose world average is 260 ms. Ratios: 28 ÷ 30 = 0.93 (good) and 190 ÷ 260 = 0.73 (good, just above the great limit). Both are healthy even though the raw numbers are very different. A 190 ms result looks poor in isolation but is faster than the typical visitor.

Example 3: Estimating a floor for a call

A call between London and Sydney crosses roughly 17,000 km. Fibre minimum RTT is 170 ms, and real paths bring this to about 250 to 300 ms. One-way delay of 125 to 150 ms is at the limit of comfortable conversation, so some talking-over is expected whatever the connection. Pushing for 100 ms is impossible; stable 270 ms is a good result.

How to judge your own result

1. Identify the destination and its approximate distance. Use 1 ms per 100 km as the floor and double it for a realistic expectation.
2. Take several samples. Use the average, best and worst, not one reading.
3. Compare with a reference. A world average for the same destination, or a nearby server that behaves as expected.
4. Check stability. A worst pass much higher than the average signals jitter or congestion.
5. Test wired. If cable is much better than Wi-Fi, the problem is local.

Common mistakes and myths

• "Anything above 50 ms is bad." Only for nearby servers. Distance matters.
• Chasing the lowest possible number. Beyond a threshold for the activity, improvements are not perceptible.
• Comparing ICMP ping with an HTTPS-based test. They measure differently; see ping vs HTTPS latency.
• Judging by the average alone. Spikes ruin calls and games even when the average is low.
• Blaming the connection for a distant server. The server's location may be the whole explanation.

What to do next

Run the latency test against destinations at several distances and see how each compares with the world average. If you are well above the norm on nearby destinations, follow troubleshoot high latency. For gaming specifically, read latency for gaming, and for calls latency for video calls and VoIP.