ASCII Table

Control characters (0–31 and 127)

The first 32 codes and code 127 are non-printing control characters. They were designed for teleprinters and serial terminals: ring the bell, advance the paper, return the carriage. Most are historical curiosities today, but a handful remain essential.

The ones that still matter

• 9 (HT) — tab, still used for indentation and TSV files.
• 10 (LF) — line feed, the newline on Unix, macOS and inside HTTP and most network protocols.
• 13 (CR) — carriage return. Windows uses CR+LF, which is why files moved between systems show stray ^M characters or collapse into a single line.
• 27 (ESC) — starts every ANSI escape sequence, which is how terminals do colour and cursor movement.
• 0 (NUL) — the string terminator in C. Injecting it into text handled by C libraries is a classic truncation attack.
• 7 (BEL) — still makes most terminals beep.

Printable characters (32–126)

These 95 characters are the ones you can type and see. Every keyboard, font and protocol supports them, which is why they form the safe common denominator for identifiers, URLs and file formats.

Patterns worth memorising

The ASCII layout is not arbitrary. Several deliberate design choices make character manipulation cheap in code:

• Digits start at 48. '7' - '0' gives the numeric value 7. This is why char - 48 appears in so much parsing code.
• Case differs by exactly 32. Uppercase A is 65, lowercase a is 97. Since 32 is 2⁵, flipping bit 5 toggles case: c ^ 32. c | 32 forces lowercase and c & ~32 forces uppercase.
• Letters are contiguous. A–Z run 65–90 and a–z run 97–122 with no gaps, so range checks work. EBCDIC, the IBM alternative, did not have this property and was far more painful to program against.
• Control characters come first. Any byte under 32 is non-printing, so a single comparison filters them.

A short history

ASCII — the American Standard Code for Information Interchange — was published by the American Standards Association in 1963 and revised in 1967 and 1986. Before it, every manufacturer used its own encoding, and moving data between an IBM machine and anything else required a translation table.

The committee chose 7 bits, giving 128 codes. Seven was a compromise: six bits could not fit both cases plus digits and punctuation, and eight would have wasted expensive memory and transmission time. The eighth bit was commonly used as a parity check on serial lines.

That decision echoes today. UTF-8 was designed so that the 128 ASCII codes encode as single bytes with identical values, which means every ASCII file is already a valid UTF-8 file. Sixty years on, the table below is still the foundation of Unicode and therefore of essentially all digital text.

Extended ASCII and why it is ambiguous

“Extended ASCII” refers to codes 128–255, and the phrase is misleading because there is no single extended ASCII. Dozens of incompatible extensions filled that range:

The same byte meant different characters in each, which is the root cause of mojibake — text turned to garbage by being decoded with the wrong table. Unicode and UTF-8 exist to end this, and a page that declares <meta charset="utf-8"> never has the problem. Our character encoding guide traces the whole story.

Frequently Asked Questions