head

What it does

head prints the beginning of one or more files — by default the first 10 lines. It's the natural counterpart to tail, and it's especially handy for quickly previewing large files, checking headers or column names in data files, or trimming down the output of a long-running command in a pipeline.

Beginner examples

• head file.txt — print the first 10 lines
• head -n 5 file.txt — print the first 5 lines
• head -n 1 file.csv — print just the header row of a CSV
• head file1.txt file2.txt — print the first 10 lines of each, with filename labels

head -n 20 access.log

Advanced examples

• Preview raw bytes rather than lines, useful for binary files: head -c 64 image.png
• Print everything except the last N lines (GNU head): head -n -10 file.txt
• Suppress filename headers when checking multiple files: head -q -n 3 *.txt
• Combine with tail to grab a specific middle chunk: head -n 50 file.txt | tail -n 10 (lines 41-50)
• Quickly check the shape of a large CSV before processing it: head -n 1 data.csv

head -n 50 report.csv | tail -n 10

Common mistakes

• Assuming head reads the whole file first — it doesn't need to for line-based output, but be aware -c byte counts are exact regardless of line boundaries and can cut a line mid-character for multi-byte encodings.
• Forgetting the default is 10 lines and being confused by unexpectedly short output.
• Using head -n -N on macOS/BSD head, which doesn't support the negative-count GNU extension the same way; check man head on non-Linux systems.
• Reaching for head when you actually want the end of a file — that's tail's job.

Tips

• head -n 1 file is a quick way to grab just a CSV or TSV header row for scripting.
• Pipe verbose command output through head to avoid scrolling past what you need, e.g. find / -name "*.conf" | head -n 20.
• Combine head and tail to extract an arbitrary line range from a file without a full text editor.

Best practices

• Use head instead of opening large files in an editor just to check their format or first few records — it's far faster and doesn't risk accidentally modifying a huge file.
• In scripts, prefer explicit -n counts over relying on the default 10, so behavior stays clear to future readers.
• When working with binary data, use -c (bytes) rather than -n (lines), since binary files often have no meaningful line structure.

Real-world use cases

• Quickly checking the column headers of a large CSV export before writing a processing script.
• Sampling the first few log entries after a deployment to confirm logging is working: head -n 20 app.log.
• Verifying a downloaded file's type by inspecting its first bytes: head -c 16 file.bin | xxd.

Common interview questions

• How would you print the first 20 lines of a file? head -n 20 file.txt.
• What's the difference between head and tail? head prints the beginning of a file (default first 10 lines); tail prints the end (default last 10 lines).
• How do you get just the header row of a CSV file? head -n 1 file.csv, which prints only the first line.