watch Command in Linux: Run Commands Repeatedly

Quick take: The watch command re-runs a command at a set interval and shows its latest output full-screen. watch df -h refreshes every 2 seconds; -n 1 sets a 1-second interval and -d highlights what changed.

Introduction

The watch command repeatedly runs another command and displays its output, refreshing the screen each time. It turns any one-shot command into a live monitor — perfect for watching disk usage, queue lengths, file counts, or service status change in real time without writing a loop.

Syntax

The basic syntax of the watch command is:

watch [OPTIONS] COMMAND

Common Options and Parameters

The most useful options and parameters for the watch command:

Practical Examples

Real watch commands you can run today:

# Watch disk usage update
watch df -h
# Refresh every second
watch -n 1 free -h
# Highlight what changes
watch -d 'ls -l /var/spool/mail'
# Monitor a process count
watch -n 2 'ps aux | grep -c nginx'
# Watch a directory's file count
watch 'ls /incoming | wc -l'
# Exit when output changes
watch -g ls /tmp/ready

Tips and Best Practices

• Quote the command when it contains pipes or globs (watch 'ls | wc -l') so watch runs the whole pipeline, not just the first part.
• Use -d to highlight changes, which makes it easy to spot what moved between refreshes.
• -g turns watch into a simple wait-for-change trigger, exiting the moment the output differs.

Final Thoughts

watch converts any command into a live dashboard, re-running it on an interval and refreshing the screen. Set the cadence with -n, highlight changes with -d, and quote piped commands so the whole pipeline runs. It is the quickest way to monitor something changing without scripting a loop, complementing top and tail -f.