cut Command in Linux: Extract Fields, Bytes & Practical Examples

Deepak Prasad
Tested on Ubuntu 25.04 (Plucky Puffin)
Package coreutils-from-uutils
cut (uutils coreutils) 0.2.2
Applies to Ubuntu
Privilege sudo or root
Man page cut(1)
Scope cut prints selected bytes, characters, or delimiter-separated fields from each line of a file or pipeline. It is the fast first choice for usernames from /etc/passwd, CSV columns, and fixed-width snippets in shell.
Related guides grep
Linux commands

cut — quick reference

Field mode

Extract columns separated by a delimiter — default is tab; colon is common in /etc/passwd.

When to use Command
First field with default tab delimiter cut -f 1 file
First field with colon delimiter cut -d ':' -f 1 /etc/passwd
Multiple fields (non-contiguous) cut -d ':' -f 1,3 /etc/passwd
Range of fields cut -d ':' -f 1-3 /etc/passwd
From field N to end of line cut -d ':' -f 3- /etc/passwd
Only lines that contain the delimiter cut -d ':' -f 1 -s file
Split on any run of spaces/tabs (uutils -w) cut -w -f 1,3 file

Character and byte mode

Fixed-position extraction when data is not delimiter-separated.

When to use Command
Characters 1 through 5 on each line cut -c 1-5 file
Bytes 1 through 10 on each line cut -b 1-10 file
From character 5 to end cut -c 5- file
From start through character 10 cut -c -10 file
Non-contiguous character positions cut -c 1-5,10-15 file

Output control

When to use Command
Print everything except selected fields cut --complement -d ':' -f 2 file
Replace output delimiter (e.g. tab → space) cut -d ':' -f 1,3 --output-delimiter=' ' /etc/passwd
NUL-terminated records cut -z -f 1 file

Pipelines and metadata

When to use Command
Extract from command output ps aux | cut -c 1-20
Chain with grep ps aux | grep root | cut -c 1-30
Show built-in usage cut --help
Show package version cut --version

cut — command syntax

Synopsis from cut --help on Ubuntu 25.04 (uutils cut 0.2.2):

text
cut OPTION... [FILE]...

Each invocation must pick a mode (-b, -c, or -f) and a list of positions or fields. With no FILE, cut reads stdin. Delimiter for -f defaults to tab unless -d or -w is set.

cut — command examples

Essential Extract usernames from /etc/passwd

The classic field cut: colon-separated fields, keep field 1.

Run the command:

bash
cut -d ':' -f 1 /etc/passwd | head -3

Sample output:

output
root
daemon
bin

Field 1 is the login name. Field 3 is UID, field 6 is home directory — same delimiter throughout the file.

Essential Select several fields from one line

Comma-separated field numbers print in the order you list them.

Run the command:

bash
cut -d ':' -f 1,3 /etc/passwd | head -2

Sample output:

output
root:0
daemon:1

Here you get username and UID without the password placeholder in field 2.

Essential From field 3 to end of line

A trailing hyphen means "through the last field" — useful when trailing fields vary in count.

Run the command:

bash
cut -d ':' -f 3- /etc/passwd | head -2

Sample output:

output
0:0:root:/root:/bin/bash
1:1:daemon:/usr/sbin:/usr/sbin/nologin

Everything from UID onward is kept. Pair with --output-delimiter when you need spaces instead of colons in the result.

Common Invert selection with --complement

Print all fields except the ones you name — handy to drop a middle column.

Run the command:

bash
cut --complement -d ':' -f 2 /etc/passwd | head -2

Sample output:

output
root:0:0:root:/root:/bin/bash
daemon:1:1:daemon:/usr/sbin:/usr/sbin/nologin

Field 2 (the x password placeholder) is removed; other colons stay.

Common -s skips lines without the delimiter

By default, cut prints whole lines that lack the delimiter. -s suppresses them — useful on messy logs.

Run the command:

bash
cut -d ':' -f 1 -s /etc/shadow | head -3

Sample output (requires read access to /etc/shadow):

text
root
daemon
bin

Only lines containing : are printed. On files where some lines are free text, -s keeps output clean.

Common Change the separator in output

--output-delimiter replaces the field separator in the printed columns.

Run the command:

bash
cut -d ':' -f 1,3 --output-delimiter=' ' /etc/passwd | head -2

Sample output:

output
root 0
daemon 1

Downstream tools that dislike colons (or tabs) can consume space-separated values directly.

Common Fixed character positions with -c

When columns are defined by width, not delimiters, use character mode.

Run the command:

bash
printf 'abcdef\nghijkl\n' > /tmp/cut.txt
cut -c 1-3 /tmp/cut.txt

Sample output:

output
abc
ghi

-b selects bytes (same result on ASCII); -c counts Unicode characters, which matters for multibyte text.

Common Trim wide command output in a pipeline

cut fits between filters when you only need the leftmost part of aligned output.

Run the command:

Find the blocking PID with ps aux piped to grep; the ps command explains BSD versus UNIX options and filtering command lines.

bash
ps aux | head -3 | cut -c 1-20

Sample output:

output
USER         PID %CP
root           1  0.
root           2  0.

For delimiter-separated command output (e.g. df), prefer -f with the right -d — or awk when spaces vary in width.

Advanced Whitespace-separated fields with -w (uutils)

On Ubuntu 25.04's uutils cut, -w treats any run of spaces or tabs as one separator — closer to how humans read columns.

Run the command:

bash
printf 'one two  three\n' | cut -w -f 1,3

Sample output:

output
one	three

Output delimiter defaults to tab unless you set --output-delimiter. GNU coreutils on other distros may not ship -w — check cut --help on the target host.

cut — when to use / when not

Use cut when Use something else when
  • Columns are separated by a single-character delimiter (tab, colon, comma)
  • You need one or more field numbers quickly in a shell pipeline
  • Data uses fixed character or byte positions
  • You want complement, output delimiter, or serialize-style field drops without a full language
  • Columns are separated by variable spaces → awk or cut -w where available
  • You need conditional logic, calculations, or pattern matching → awk
  • You need to merge parallel files into wider rows → paste
  • You need in-place editing → sed

cut vs awk

cut awk
Learning curve Small — one delimiter, field list Larger — patterns and expressions
Delimiter Single character (-d) or -w whitespace Any pattern in FS
Logic No conditionals Filters, math, keyed joins
Speed on huge files Very fast, minimal Fast, more flexible

Use cut for simple column pulls; reach for awk when rules get conditional.

cut — interview corner

What is the cut command used for?

cut prints selected bytes (-b), characters (-c), or fields (-f) from each input line. It is built for simple column extraction in scripts.

bash
cut -d ':' -f 1 /etc/passwd

A strong answer is:

"cut extracts fixed positions or delimiter-separated fields from each line — I use it for passwd columns and quick pipeline trimming."

What is the default field delimiter in cut?

The default is tab. For colon files like /etc/passwd, set -d ':'.

Lines without the delimiter are printed whole unless you add -s (suppress).

A strong answer is:

"Tab by default; I set -d for colons or commas. -s drops lines that lack the delimiter."

When would you use awk instead of cut?

Use awk when delimiters are irregular (multiple spaces), you need conditions ($3 > 100), or you are matching patterns. cut wins when the format is stable and the task is "give me fields 1 and 3."

A strong answer is:

"awk for variable spacing, conditions, and patterns; cut for simple single-character delimiters and fixed field lists."

What does cut --complement do?

It inverts the field list — you get every field except the ones you named.

bash
cut --complement -d ':' -f 2 /etc/passwd

drops field 2 (the x password token) and keeps the rest.

A strong answer is:

"--complement prints all fields except those listed — useful to drop one column without renumbering the ones you want."

How do you specify a range of fields in cut?

Use N-M for inclusive range, N- from N to end, -M from start to M. Combine with commas: -f 1,3-5.

bash
cut -d ':' -f 1-3 /etc/passwd

A strong answer is:

"Ranges use hyphens — 1-3 or 3- for rest of line; commas combine multiple selections."

Troubleshooting

Symptom Likely cause Fix
Whole lines printed when you expected one field Line has no delimiter; -s not set Add -s or fix delimiter with -d
Empty columns with -d ' ' Multiple spaces count as multiple delimiters (without -w) Use cut -w on uutils or switch to awk
cut: you must specify a list of bytes, characters, or fields Missing -f, -c, or -b Add mode and list, e.g. -f 1
Wrong characters on UTF-8 text Byte vs character mode Prefer -c for characters, -b for raw bytes
--complement not recognized Very old coreutils Upgrade or select fields positively without complement
Rohan Timalsina

is a technical writer and Linux enthusiast who writes practical guides on Linux commands and system administration. He focuses on simplifying complex topics through clear explanations.

Expertise
  • Linux
  • HTML5
  • JavaScript
  • Web Design
  • Front-end Web Development

Related Articles

linux

Bash Shell Scripting for Linux Administrators

Learn Bash shell scripting with variables, arguments, conditions, loops, file input, exit codes, debugging, …

13 min readlinux

pkill -P: Kill Child Processes by Parent PID in Linux

Use pkill -P PARENT_PID . to send SIGTERM to every direct child of a parent process, list them with ps --ppid, …

5 min readcheatsheet

paste Command in Linux: Merge Lines, Delimiters & Practical Examples

paste merges corresponding lines from one or more files into wider rows — tab-separated by default, or with …

7 min read