# Mean Time to Recovery (MTTR)

> The average time it takes to fix a service that broke. Lower is better.

**Field:** Infrastructure  
**Category:** Reliability  
**Also known as:** MTTR, Mean Time to Repair, Mean Time to Restore, Time to Restore Service

## Think of it like

It is like timing how long it takes to fix a flat tire. Then you take the average of many.

## Definition

Mean time to recovery is the average time from when a service breaks to when it works again. You add up how long each incident lasted. Then you divide by the number of incidents. The R can mean recovery, repair or restore, so say which you mean. A falling MTTR shows a team is getting faster at fixing things.

## Formula

```
MTTR = total time down ÷ number of incidents
```

## Example

Jack's team had three incidents this quarter. They lasted 20, 40 and 60 minutes, so the MTTR is 40 minutes.

## Why it matters

Failures will happen. How fast you recover is often easier to improve than how often you fail.

## Common mistake

Comparing MTTR between teams that count the start and end in different ways. Agree on both first.

## Three principles

- **An average.** Total time down over number of incidents.
- **Say which R.** Recovery, repair or restore differ.
- **Watch the trend.** Falling over time is the goal.

## Related terms

- [Incident (IT)](https://learn.benibix.com/term/infrastructure/incident)
- [On-Call](https://learn.benibix.com/term/infrastructure/on-call)
- [Runbook](https://learn.benibix.com/term/infrastructure/runbook)
- [High Availability (HA)](https://learn.benibix.com/term/infrastructure/high-availability)
- [Mean Time to Detect (MTTD)](https://learn.benibix.com/term/infrastructure/mean-time-to-detect)
- [Mean Time Between Failures (MTBF)](https://learn.benibix.com/term/infrastructure/mean-time-between-failures)

## Told apart from

- [MTBF vs MTTR](https://learn.benibix.com/compare/mtbf-vs-mttr): People read the two numbers as one score for reliability, and they measure different things. MTBF is how long a system runs, on average, before it breaks. MTTR is how long it takes to bring it back. A service can break often but recover in seconds, or break rarely and stay down for a day.

![A timeline from left to right. Something breaks. An alert fires and pages the person on call. They open the runbook and fix it, and the service is back. The span from break to back is marked as time to recover. After that, a postmortem is written, and its actions feed back to the start.](https://learn.benibix.com/diagrams/inf-incident-timeline.svg)

---

Source: https://learn.benibix.com/term/infrastructure/mean-time-to-recovery
Author: Behnia Zarei (a developer in Vienna, Austria, who builds and teaches data, AI and business intelligence at BENIBIX)
Publisher: BENIBIX Learn (https://learn.benibix.com/)
Last updated: 2026-11-19
Licence: CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). Quote freely with attribution.
