The transliteration of the Arabic script to Latin is complex in part because Arabic is a non-vocalized script, that is, vowels are not normally written. A pure transliteration should also follow this convention, so that the word قطر, for example, is transliterated as qṭr. In practise, however, this renders the Romanized form meaningless to an untrained reader, so vowels are often inserted where appropriate. The table below only shows the “pure” Latin equivalents without vowels.

The table compares six transliteration standards, as well as the ad-hoc system which is often used in informal electronic contexts such as chatrooms, email, and on social media sites.
The ISO 233 standard shown here was established in 1984. A simplified version was created in 1993 for the purpose of library indexing, but this is not included here.
The American Library Association - Library of Congress (ALA-LC) created an Arabic romanization standard in 1997 for indexing and cataloguing purposes, although it is now used in other academic and journalistic contexts also.
ArabTeX is a software package for typesetting texts which contain both the Roman and the Arabic scripts. These texts often include transliterations of Arabic portions, so a standard was created to address this.
The United Nations Group of Experts on Geographical Names (UNGEGN) uses a romanization system sometimes referred to as the Amended Beirut System, Variant A.
The Institut Géographique National (IGN) has created a standard more conformant with French orthography, which is commonly used in French-speaking areas such as Maghreb and Lebanon. It is also referred to as the Amended Beirut System, Variant B.
The Deutsches Institut für Normung adopted the DIN 31635 standard in 1982. Its popularity owes in part to its elegance, assigning one Latin character to one Arabic letter.
In informal contexts where users only have access to a standard Latin keyboard, an ad-hoc romanization system is applied. Latin equivalents are chosen based either on phonetic or visual similarity to their Arabic counterparts. This system has not been standardized, so different users may apply different Arabic-Latin correlations. The table shows only the most commonly used values.

For more information about these standards, and an explanation of the complex rules governing their application, please download the  ALA-LC standard from the Library of Congress or the  UN Working Group Report on Arabic Romanization, both in PDF format, or refer to the  Wikipedia page on Arabic romanization.

ArabicISO 233ALA-LCArabTeXUNGEGNIGN 1973DIN 31635Common informal
  ء ˈ, ˌ   2
  ا ʾ ā a ā a, e, é, è ā a / e / é
  ب b b b b b b b
  ت t t t t t t t
  ث th _t th th s / th
  ج ǧ j g j dj, j ǧ j / g / dj
  ح .h 7
  خ kh _h kh kh kh / 7' / 5
  د d d d d d d d
  ذ dh _d dh dh z / dh / th
  ر r r r r r r r
  ز z z z z z z z
  س s s s s s, ss s s
  ش š sh s sh ch š sh / ch
  ص .s ş ṣ, ç s / 9
  ض .d d / 9'
  ط .t ţ t / 6
  ظ .z z / dh / 6’
  ع ʿ ʻ ` ʻ ’, aa ʿ 3
  غ g gh .g gh gh ġ gh / 3’ / 4
  ف f f f f f f f
  ق q q q q q, g, gu q 2 / g / q / 8
  ك k k k k k k k
  ل l l l l l l l
  م m m m m m m m
  ن n n n n n, ne n n
  ه h h h h h h h
  و w; ū w; ū w w ou w w; o; ou / u / oo
  ي y; ī y; ī y y i, ï, y y y; i / ee; ei / ai
  آ ʾâ ā, ʼā 'A ā â, ê ʾā 2a / aa
  ة h; t T h; t h; t h, t a / e(h); at / at
  ى à á _A y y ā a; i / y
  ال ʾal al- al- al- al- al- el