Calculate round-trip time (RTT) from network distance and signal medium, or from one-way delays, and solve for a missing one-way latency.

Uses the actual cable route length, which is usually longer than the straight-line distance.

Typical signal speeds; fiber carries light at about two-thirds of c.

Other round-trip delays
Blank delay fields mean 0. Enter totals for both directions, added once.

Related Calculators

RTT Formula

Round-trip time (RTT) is the total elapsed time for data to travel from a client to a server and back again. In networking, it is one of the clearest measures of connection responsiveness because it captures the full request-and-response path rather than only a single direction.

RTT = Dcto s + Dsto c

Choose RTT or a directional delay, then enter the other known values. For compatible measurements of the same exchange and conditions, RTT = forward delay + reverse delay + any remote turnaround not already included. The displayed two-term formulas assume no additional turnaround. Enter it separately only when excluded from both directional inputs; blank means zero. Independently measured one-way averages under different conditions do not generally add to a triggered-response RTT.

Dcto s = RTT - Dsto c
Dsto c = RTT - Dcto s

Variable Definitions

RTT
Total round-trip time for a packet, request, or message to go out and return.
\\(\overline{D}_{c\to s}\\)
Average one-way latency from the client to the server.
\\(\overline{D}_{s\to c}\\)
Average one-way latency from the server back to the client.

How to Use the RTT Calculator

  1. Choose RTT from distance (the default; pick a signal medium such as fiber), RTT from one-way delays, Client → Server delay, or Server → Client delay.
  2. Enter the two known compatible values, and remote turnaround only if not already included.
  3. Select a unit for each input; the calculator converts units automatically.
  4. Click calculate to get the missing value instantly.

If your network path is symmetrical, the two one-way delays may be similar. In that case, RTT is often close to twice the one-way latency.

D โ‰ˆ (RTT) / (2)

That shortcut is only an approximation. Real networks often have different delays in each direction due to routing differences, congestion, wireless conditions, or processing overhead.

Averaging One-Way Delay

Latency usually changes from packet to packet, so using an average is more informative than relying on a single measurement. If you collect multiple samples in each direction, the directional average can be calculated as follows:

Dcto s = (ฮฃแตขโ‚Œโ‚โฟ Dcto s,i) / (n)
Dsto c = (ฮฃแตขโ‚Œโ‚โฟ Dsto c,i) / (n)

Directional averages describe their measured samples. They estimate RTT only with compatible timing boundaries, packet populations and conditions; an average does not remove systematic timing error or unaccounted remote turnaround.

Unit Conversion Notes

The directional calculator supports milliseconds, seconds, microseconds, and nanoseconds, and converts them to a common unit before adding or subtracting.

1 s = 1000 ms = 10โถ ฮผ s = 10โน ns

For most application and network troubleshooting scenarios, milliseconds are the most practical unit because they are easy to compare and interpret.

Example

If the average latency from client to server is 7.354 ms and the average latency from server to client is 4.568 ms, then the RTT is:

RTT = 7.354 ms + 4.568 ms = 11.922 ms

If the RTT is already known and you want to recover the missing forward-path delay, subtract the reverse-path delay from the RTT:

Dcto s = 11.922 ms - 4.568 ms = 7.354 ms

Distance-Based Estimate

The separate propagation mode uses estimated RTT = 2 x one-way path length / propagation speed + total round-trip processing delay + total round-trip queueing delay. It assumes the same path length and speed in both directions and adds each entered delay once. Use actual cable or route length. Remote turnaround can be included in processing only if not counted elsewhere. Omitted serialization and other delays are excluded, so this is not a network measurement. Speed presets are illustrative assumptions; vacuum light speed is exactly 299,792,458 m/s.

Why RTT Matters

  • Web applications: Higher RTT can make pages, APIs, and interactive dashboards feel slower even when bandwidth is adequate.
  • Online gaming: Lower RTT generally improves responsiveness and reduces input delay.
  • Voice and video calls: Excessive RTT can make conversations feel delayed and less natural.
  • Remote desktop and cloud apps: RTT directly affects how quickly clicks, keystrokes, and screen updates are reflected.
  • Network diagnostics: Comparing RTT values across routes, servers, or time periods helps identify performance changes.

What Can Increase RTT?

  • Long physical distance between client and server
  • Congested routers or network queues
  • Wireless interference or unstable signal quality
  • VPN tunnels, proxies, or additional network hops
  • Server processing delays before sending a response
  • Packet loss and retransmissions
  • Asymmetric routing where the return path differs from the outbound path

Important Interpretation Notes

  • RTT is a round-trip measurement, not a pure one-way delay measurement.
  • Two connections with the same bandwidth can feel very different if their RTT values differ significantly.
  • If you solve for a one-way delay and get a negative value, the inputs are inconsistent and should be checked.
  • One-way measurements typically require more specialized timing or synchronized clocks, while RTT is commonly easier to observe directly.