廟宇幾時建：由宋元到十八世紀的建立時序

1. Both kinds of date are upper bounds

When the chronicles record that a temple “had already been built” or “was completed”, the building existed by that year; the year work began, the year the ridge beam was raised, or the year of consecration is usually unrecorded. The Antiquities Advisory Board list states the uncertainty even more plainly, typically as “possibly built in or before” a year. Neither wording can be presented as a foundation year.

This dossier therefore makes no claim about “the year Hong Kong’s oldest temple was founded”. It states only when a temple first appears in the documentary record. The gap between the two can be decades or longer, and unless a stone inscription, bell casting or land deed can be quoted verbatim, an upper bound should not be written as a definite year.

2. Song and Yuan: the sea came before the town

The chronicles record that around 1266 Lam To-yee, from Putian, funded the building of the Lam clan’s Tai Ku Temple at Pak Fat Tong in what is now Tai Miu Wan, Sai Kung. The bay is the same one where Tin Hau Temple at Joss House Bay stands today, which places this earliest layer of religious building beside a sea route rather than beside a market town.

In 1274 the salt official Yim Yik-cheung of Kwun Fu Cheung visited the same bay and left an inscription on rock, showing that officials travelled there and that salt production was administered nearby. Temples overlapping with salt fields and shipping lanes is the most striking feature of Hong Kong’s earliest religious architecture: the shrine did not follow the town; it arrived with maritime activity.

3. Between Ming and Qing: Kwan Tai and Che Kung enter the villages

The chronicles record that the Kwan Tai Temple at Tai O was completed between 1488 and 1505, adding that it is “the earliest Kwan Tai temple in the territory for which documentary evidence survives”. That is a claim about documentary survival, not about being first in absolute terms — a distinction worth keeping whenever a temple is called “the oldest”.

By 1644 the Che Kung Temple at Ho Chung, Sai Kung, had been built; the chronicles describe it as the oldest surviving Che Kung temple in Hong Kong. The Ming dynasty fell that same year, and the coastal evacuations and return that followed reshuffled village populations. In that setting a temple was more than a shrine: it was where villages settled disputes, held meetings and stored communal property.

In 1692 the Kwan Tai Temple at Mong Tseng Wai, Yuen Long, had already been completed; in 1699 the Yeung Hau Temple at Tai O was completed. One stood inside a walled village, the other beside a fishing market, showing that by the end of the seventeenth century religious building had split into two lines: protective shrines within villages, and temples serving mobile populations at coastal markets.

4. The eighteenth century: temples follow markets and sea routes

By 1730 the Hau Wong Temple at Kowloon City had been built. It stands at the junction of Junction Road and Tung Tau Tsuen Road today, facing what was the old shoreline of Kowloon Bay; Hau Wong devotion circulated through village networks in the New Territories and Kowloon, and such temples often became nodes for activity across several villages.

In 1747 the Tin Hau Temple at Causeway Bay had already been built. Tin Hau devotion is tied directly to work at sea, and in the mid-eighteenth century Causeway Bay was still a bay, with the temple on its shore; the building now looks inland because the city grew around it, but its siting only makes sense against the coastline of the time.

In 1773 the Hung Shing Temple on Ap Lei Chau was completed. Hung Shing devotion was widespread among fishermen and boat people around the Pearl River estuary, and Ap Lei Chau guards the channel off Aberdeen. From the Kowloon peninsula to the southern shore of Hong Kong Island, these three temples form a line along coast and market rather than an even spread across the territory.

5. How to read the sequence, and what it cannot tell you

This sequence records only when a temple first appears in the documentary record and when the official list considers it possibly to have existed. It cannot be used to rank temples by age. There are three reasons: whether a document mentions a temple depends on whether its compiler noticed it; a temple may be rebuilt on the same site, so construction and documentation are separate questions; and some temples share a name while having moved, so a shared name does not mean a shared building.

A practical trap in reading: the same district can hold two or three temples sharing a name — Tin Hau temples exist on Hong Kong Island, in Kowloon and on the outlying islands — with different addresses and different building histories. Merging them by name alone would present buildings of different periods as one. Each entry in this dossier therefore names its district rather than using a bare temple name for the whole territory.

One more point: every year cited here comes from the official wording of the chronicles or the Antiquities Advisory Board list. None is inferred from architectural style, timber framing or inscriptions. If a stone inscription, bell casting or deed can later be quoted verbatim, the upper bound may move earlier; until then only the bound is stated.