Urdu Script Overview
The word urdū itself is the fastest lesson in how this script behaves. All four of its letters — ا ر د و — refuse to join to the letter that follows them, so the word is written in four separate pieces. Roughly a third of the alphabet does this. Knowing which letters break is most of what makes handwritten Urdu readable.
Short vowels aren't written
Urdu is an abjad: only consonants and long vowels appear. کتب could be kutub or katab — you supply the short vowels from knowing the word. Diacritics exist but are almost always omitted.
Dots do the heavy lifting
Five letters — ب پ ت ٹ ث — share one skeleton and differ only in dots and marks. Same for ج چ ح خ and ع غ. Learn the skeletons first, dots second.
One sound, several spellings
s is س ص ث. z is ز ذ ض ظ. t is ت ط. They sound identical in Urdu; which one a word takes is inherited from Arabic and simply has to be memorised.
ھ is a modifier, not a letter
Do-chashmī he adds a puff of breath to the consonant before it: ک → کھ, ب → بھ. It never starts a word and never stands alone.
Notes on this table
Order. Standard Urdu dictionary order (حروفِ تہجی), beginning with alif madda as requested. Counts of 38, 39 and 40 all circulate in print — the disagreement is over whether آ, ء and ۃ count as letters in their own right or as variants. Nothing here is missing either way.
Forms. The kashida stroke ـ in the initial, medial and final cells stands for whatever letter joins on that side; it is not part of the letter. Cells hatched grey mean the form doesn't exist, because the letter never joins to its left.
Typeface. All Urdu here is set in Noto Nastaliq Urdu — the face iOS and Google use to render Urdu, and the reason صبح looks the way it does on your phone. Nastaliq is the style Urdu is actually published in: each word cascades down and to the left rather than sitting on a flat line, and the jīm family ج چ ح خ gets the deep open bowl that makes it legible at a glance. It's licensed under the SIL Open Font License, and it's embedded directly in this file rather than fetched from a CDN — so the page makes no external requests and renders the same offline as online.
Reading a Nastaliq word. The baseline slopes. Start at the top right of a cluster and follow it down to the left; the letter that looks like it's sitting lowest is usually the last one in the word, not the first. Dots may be stacked well above or below the letter they belong to — assign them by counting, not by position.
Romanisation is loose ALA-LC-flavoured transliteration, chosen for readability rather than strict reversibility. Long vowels are marked with a macron (ā ī ū), retroflex consonants with a dot below (ṭ ḍ ṛ).
Compiled from the Ling App alphabet tables, the Wikipedia article on the Urdu alphabet, and Omniglot's positional-form charts, cross-checked against each other.
Notes on this table
Order. Standard Urdu dictionary order (حروفِ تہجی), beginning with alif madda. Counts of 38, 39 and 40 all circulate in print. The disagreement is over whether آ and ء count as letters in their own right or as variants, and over where ھ and ء fall in the sequence. This table follows the majority convention and notes the disagreement in the relevant rows. Nothing is missing either way.
Forms. The kashida stroke ـ in the initial, medial and final cells stands for whatever letter joins on that side. It is not part of the letter. Cells hatched grey mean the form does not exist, because the letter never joins to its left.
Typeface. All Urdu here is set in Noto Nastaliq Urdu. Nastaliq is the style Urdu is actually published in: each word cascades down and to the left rather than sitting on a flat line, and the jīm family ج چ ح خ gets the deep open bowl that makes it legible at a glance. Most fonts covering Arabic script render Urdu in Naskh instead, which is readable but not what an Urdu reader expects to see. Noto Nastaliq Urdu is licensed under the SIL Open Font License 1.1.
Reading a Nastaliq word. The baseline slopes. Start at the top right of a cluster and follow it down to the left. The letter that looks like it is sitting lowest is usually the last one in the word, not the first. Dots may be stacked well above or below the letter they belong to, so assign them by counting rather than by position.
Romanisation. The transliterations here loosely follow ALA-LC, the standard American and Canadian libraries use to catalogue Urdu-language material. Long vowels take a macron (ā ī ū) and retroflex consonants take a subscript dot (ṭ ḍ ṛ), which is how ٹ is distinguished from ت. I have not followed it strictly. Full ALA-LC preserves distinctions that matter only if you need to reconstruct the original spelling from the romanisation, and what is here is meant to be read aloud instead.
Sources and method
I built this as study material for myself, and I am putting it online in case other learners find it useful.
The table cross-references several existing references rather than reproducing any one of them. The letter inventory, dictionary order, and positional-form data follow the Wikipedia entry on the Urdu alphabet, checked against positional-form charts. Letter names and romanisation conventions follow the tables published by the Ling App.
Three errors in the source tables have been corrected here: طوطا glossed as "broken" rather than "parrot", ژالہ باری transliterated as "Yalabari", and ہ and ن each listed twice, as though their alternate names denoted separate letters.
The sample vocabulary, English glosses, articulatory descriptions, IPA values, and everything in the notes column are not drawn from those sources. I wrote them for this table. All of it was then reviewed by my husband, a native Urdu speaker, who corrected the example words, the glosses, and the sound descriptions.
This is a learner's reference rather than a philological edition. Corrections are welcome.
References
"Urdu alphabet." Wikipedia. Accessed 25 July 2026. en.wikipedia.org/wiki/Urdu_alphabet
Ling App. "Know All The 36 Useful Parts Of The Urdu Alphabet." Medium. Accessed 25 July 2026.
Positional-form charts (isolated, initial, medial, final) consulted in facsimile. Publisher not identified.
Other resources
Two textbooks I have found useful alongside classes:
Matthews, D., & Dalvi, M. K. (2017). Complete Urdu. Teach Yourself. In English.
Oesterheld, C., & Tahir, A. M. (2016). Urdu für Anfänger. Buske. In German.