The Urdu Alphabet: A Complete Beginner's Guide

The Urdu alphabet has 38 to 40 letters (depending on how you count), runs right to left, and is written in a flowing calligraphic style called Nastaliq. It looks intimidating. It isn't, once you know one secret: most Urdu letters are the same basic shape with different dots.

This guide covers every letter, the sounds that don't exist in English, and the one thing no chart can teach you.

What script does Urdu use?

Urdu is written in a modified Perso-Arabic script. It started from the Arabic alphabet, picked up four Persian additions (پ چ ژ گ), then added its own letters for sounds specific to the subcontinent, like the retroflex consonants ٹ, ڈ, and ڑ.

Three things to know before the letter chart:

• It's written right to left. Numbers within a text still run left to right.
• It's cursive by design. Letters connect to their neighbors, in print and handwriting alike. There is no "block letter" version.
• The standard style is Nastaliq. Unlike Arabic's horizontal Naskh style, Nastaliq flows diagonally: words start high and cascade down to the baseline. It's beautiful, and it's what you'll see in newspapers, books, and signs across Pakistan and India.

How many letters are in the Urdu alphabet?

You'll see 38, 39, and 40, and all of them are defensible. The core alphabet has 38 letters. Add hamza (ء) and you get 39. Some charts also list do chashmi he (ھ) as a separate letter, giving 40. The chart below lists 39. Don't stress over the count; every chart teaches the same sounds.

The Urdu alphabet chart

Read the letters right to left within each group. Transliterations are approximate; several letters share the same sound in Urdu (more on that below).

The be family and friends

Notice ب پ ت ٹ ث: one boat-shaped body, distinguished only by dots above or below. This is the pattern that makes the alphabet learnable.

The jim family

The dal and re families

Sin through ghain

Why four letters for z and three for s? They came from Arabic, where each has a distinct sound. Urdu kept the spellings of Arabic loanwords but merged the pronunciations. You only need to recognize them; producing one z sound covers all four.

Fe to the end

Letters change shape depending on position

Because the script is cursive, each letter has up to four forms: isolated, initial, medial, and final. Here's be:

This sounds like it quadruples the work. In practice it doesn't: the core shape stays recognizable, and the connected forms mostly just lose their tail. A handful of letters (ا د ڈ ذ ر ڑ ز ژ و ے) never connect to the letter after them, which is why Urdu words have small gaps inside them.

The catch: short vowels are invisible

This is the single most important thing to understand about the Urdu script. Only long vowels (ا و ی ے) are written as letters. Short vowels (a, i, u) are marked with tiny optional signs called zabar, zer, and pesh, and in real-world Urdu, nobody writes them.

So دل (dil, "heart") is spelled just d-l. The word محبت (mohabbat, "love") is spelled m-h-b-t. The reader supplies the vowels from knowing the word.

Which means: you cannot learn to read Urdu from letter charts alone. You can decode the consonant skeleton, but only hearing the word tells you it's "dil" and not "dal" or "dul". Every written word you learn needs to arrive with its sound attached. Learn words by ear first, then map the spelling onto what you already hear. Reading Urdu is recognizing words you know, not sounding out words you don't.

The retroflex letters: ٹ ڈ ڑ

These three are Urdu's own additions, marked with a small ط sign on top. Retroflex means the tongue curls back and taps the roof of the mouth. English t and d are actually closer to the retroflex ٹ and ڈ than to the soft dental ت and د, which are made with the tongue against the teeth, like in Spanish or French.

Real words: ٹماٹر (ṭamāṭar, "tomato"), ڈاکٹر (ḍākṭar, "doctor"), لڑکا (laṛkā, "boy"). The dental/retroflex contrast changes meaning, and it's nearly impossible to internalize from written descriptions. You need to hear minimal pairs from a native speaker.

Do chashmi he: the aspiration marker

Urdu distinguishes plain and aspirated consonants: b vs bh, p vs ph, k vs kh, and so on. Aspiration is a puff of breath, and it's written with ھ (do chashmī he, "two-eyed he") glued after the consonant:

Don't confuse ھ with the regular h sound, which is written ہ (choṭī he). One marks breath on the previous consonant; the other is a consonant of its own.

How to actually learn the Urdu alphabet

A plan that works:

1. Learn by shape families, not alphabetical order. Master the ب body once and you've unlocked five letters. The whole alphabet is maybe 15 base shapes plus dots.
2. Read real words from day one. Once you know ا ب ت ک م ن س ل, you can already read سلام (salām), کتاب (kitāb, "book"), and آم (ām, "mango"). Words stick; isolated glyphs don't.
3. Pair everything with native audio. Since short vowels aren't written, the sound of a word is information the page doesn't give you. Hear it, say it out loud, then read it.
4. Use spaced repetition. Letter forms and vocabulary both fade fast without review. A few minutes of SRS daily beats a weekend cramming session.

This is exactly how Hyperpolyglot approaches scripts like Urdu: you learn full real phrases with natural native audio, so the invisible vowels, the retroflex taps, and the aspiration come in through your ear while the script comes in through your eye. Built-in spaced repetition keeps both from fading. Available on iOS, Android, and Web.

The Urdu alphabet looks like 39 problems. It's really 15 shapes, some dots, and a listening habit. Start with the be family today, read your first word this week, and let audio carry the rest.

Keep reading

• The Hardest Languages to Learn
• The Language Islands Method