In the early twentieth century, China (starting with the dying Qing Empire) used Postal (Office) System Pinyin (郵政式拼音 Hanyu Pinyin: Yóuzhèngshì Pīnyīn) (unrelated to the modern Hanyu Pinyin), based on Wade-Giles (in particularly, Herbert Giles's A Chinese-English Dictionary) for postal purposes, especially for placenames on letters and stamps, and was not for universal usage. It uses some already common European names of Chinese places that override the Wade-Giles system, and incorporate some dialectal pronunciations.

The postal system was decided after the Imperial Postal Joint-Session Conference (帝國郵電聯席會議) in spring 1906 in Shanghai.

Main differences with Wade-Giles include:

• Complete lack of diacritic and accent marks.
• Chi, ch'i, and hsi (pinyin ji, qi, and xi) are represented as either tsi, tsi, and si or ki, ki, and hi depending on historic pronunciation, e.g.,
  • Peking (Pei-ching)
  • Tientsin (T'ien-chin)
  • Tsinan (Chi-nan)
• Except being the sole vowel in the syllable, the Wade-Giles u become w, e.g.,
  • Ankwo (An-kuo)
  • Chinchow (Chin-chou)
• Guangdong, Guangxi, and Fujian placenames are to be Romanized from the local dialects, such as Hakka, Cantonese, and Min (systems also obtained from Giles' A Chinese-English Dictionary).
  • Amoy (Hsia-men)
  • Swatow (Shan-t'ou)
  • Quemoy (Chin-men)
• Popular pre-existing (from 19th century of earlier) European names for place in China are to be retained, such as those of the treaty ports.
  • Canton (Kuang-chou)

See also: Romanization