Kampot and Kep Housing: Prices, Supply and Data Coverage

The largest public portal checked on 29 September 2026 returned 192 “residential” sale results for its Kampot geography. Its own type breakdown included 121 land listings, alongside 67 houses, seven villas, seven shophouses and small numbers of apartments, condominiums and other home formats. Those labels overlap, so the headline total is not a count of 192 unique homes. More importantly, a feed containing that much land cannot support a ready-made housing price benchmark.

Kep is smaller but has the same structural problem. The portal returned 58 residential results, with 38 labelled as land, 17 as houses and nine as villas, plus a handful of apartment and condominium labels. Narrowing the page to houses still leaves a mixed catalogue: conventional homes sit beside villas, house-and-land offers, unusual compounds and listings that have been online for years. An active URL therefore says much less about current supply than the word “active” suggests.

Once land-only offers, stale cards and formats that cannot be compared on a consistent basis are removed, the usable pricing core becomes much thinner. A defensible Kampot comparison can be built from four recent three-bedroom houses with stated building areas. In Kep, only two recent two-bedroom houses form an equally narrow like-for-like group. These are not estimates of total inventory; they are the portions of the public feed where property type, price and area are clear enough to support a price comparison without blending fundamentally different assets.

The second commercial catalogue adds useful evidence about format more than market size. It currently shows Amaya project villas in Kampot, including a two-bedroom unit at $320,000, while Kep has both villas and The Pearl of Kep project condominiums, including a two-bedroom unit at $245,100. Multiple units from a single development should not be read as multiple independent observations about the wider market.

Geography needs the same care as property type. The Kampot portal page spans submarkets such as Tuek Chhou and Chhuk, while the second catalogue explicitly uses province-level labels for Kampot and Kep. For a buyer considering one address, the public feed is best treated as a dated map of visible offers, not a census of either city.

Snapshot checked on 29 September 2026. Prices below are asking prices, not registered transaction prices; narrow ranges appear only where current listings are sufficiently comparable.

A portal-wide range is a poor answer to the question most buyers actually have: “Is this particular home priced sensibly against similar options?” Kampot mixes modest houses, villas, shophouses, large house-and-land properties and development stock on the same search surface. Restricting the comparison to four recent three-bedroom houses with stated building areas produces an observed asking range of $75,000 to $149,000 as of 29 September 2026. With only four records, publishing a median would add precision without adding reliability.

Other current offers sit at very different price levels because they are different products. A four-bedroom villa on the main portal was asking $375,000, while a two-bedroom Amaya project villa in the second catalogue was listed at $320,000. Folding either into the same statistic as conventional three-bedroom houses would erase the effects of development stage, land component, specification and property format.

Kep has an even thinner comparable group. Two recent two-bedroom houses with stated building areas were asking $70,000 and $75,000. A current three-bedroom villa was asking $300,000; the second catalogue shows the same villa at $300,000 with a 200 m² floor area on 875 m² of land. Seeing the same property in two catalogues is useful for cross-checking details, but it is still one market offer, not two independent price observations.

The condominium evidence is also project-specific. The second catalogue lists a two-bedroom unit at The Pearl of Kep for $245,100 with roughly 101 m² of floor area, alongside other unit types in the same development. That is evidence of current project pricing, not a defensible condominium range for Kep as a whole. The checked sources did not provide a sufficiently independent current apartment set for Kampot either.

With samples this small, the boundary around the number matters more than a single summary statistic. Four comparable Kampot houses and two Kep houses are enough to show an observed range, but not enough to call that range the typical price of either local market. A later update should rebuild the comparison from then-current offers instead of carrying the September snapshot forward as a fixed benchmark.

Every figure in this section is therefore an asking price. Negotiation, furniture, payment structure, transfer costs and the eventual registered consideration can all differ from the advertised number. The practical value of the range is diagnostic: it helps identify which differences need explaining before a buyer treats one offer as comparable with another.

Price per square metre is intuitive for apartments because the denominator is usually the internal or saleable floor area. A landed home bundles a building with a plot, and the land component can be a major part of the asking price. Dividing the same total price by floor area and by land area therefore produces two valid but fundamentally different metrics.

The current listings make the distinction visible. A $98,000 Kampot house was displayed at roughly $544 per m² of building area and $445 per m² of land; a $149,000 house showed about $931 per m² gross and $621 per m² of land. In Kep, a $75,000 house worked out at about $610 per m² of building area but $1,500 per m² of land. Those pairs are not contradictory: the denominator has changed.

There is a second problem when the source field itself is ambiguous. A current $1.6 million Kep compound in the second catalogue is a useful example: its headline describes a four-bedroom luxury villa, its property type is “Land,” and the floor-area field shows 29,231 m²—almost exactly the land size stated in the description. The text separately describes two much smaller houses on that plot. The listing is useful evidence of a field-definition problem, but it cannot support a residential $/m² calculation.

Some catalogue cards attach a very large area to a villa while the written description separately identifies a large land parcel and a much smaller residential structure. In that situation, retaining the total asking price and the separately verified building and land areas is more useful than manufacturing a neat price-per-square-metre number from an uncertain field.

For the narrow comparisons on this page, building area is used only when it is explicitly stated and consistent with the listing details. Land area remains a separate characteristic. A blended $/m² statistic across condominiums, houses, villas and land-backed compounds would be easy to calculate and difficult to interpret, so it is deliberately omitted.

Listings are sufficient to describe what sellers are advertising today. They are not a substitute for a transaction database. As of 29 September 2026, the checked public sources did not provide a dated Kampot- or Kep-specific series of registered sale prices, achieved rents or reliable resale times. That finding is about public evidence, not about whether transactions occur.

The National Institute of Statistics provides useful official housing context through the Cambodia Socio-Economic Survey 2023, but its housing section measures household and dwelling conditions, not current sale prices. Where the report uses its coastal zone, Kampot and Kep are grouped with Sihanoukville and Koh Kong. Those statistics can inform definitions and structural context; they cannot be read as a current asking or transaction benchmark for either local market.

The National Bank of Cambodia maintains the Residential Property Price Index, with the public data access page showing an update on 18 May 2026. The official material checked for this review separates Cambodia, Phnom Penh and other provinces and does not publish a standalone Kampot or Kep series. “Other provinces” therefore provides macro context only: it cannot be relabelled as a local coastal index.

Knight Frank’s Cambodia Real Estate Highlights for H2 2025 does not fill that gap. Kampot appears in the report in connection with special economic zones and industrial supply, not as a standalone residential market series, while no Kep-specific residential series was found in the checked edition. Phnom Penh residential indicators in the same report are not proxies for coastal towns.

Other missing pieces matter for investment analysis: completed-sale volume, local absorption, achieved rent, delisting outcomes and credible time-on-market data are not available as clean public series for these two locations in the sources reviewed. The page therefore stays on the side of the evidence: it describes visible asking supply and its limitations, and stops before turning that evidence into a transaction-price, yield or liquidity claim.

A useful comparable starts with a similar asset, not a similar headline price. A house in central Kampot should not automatically be benchmarked against a villa in Tuek Chhou, and a development condominium in Kep should not be compared with an older standalone house on a large plot. Source geography, development stage, building size, land size, bedroom count, condition and specification can all change the appropriate comparison set.

Recency matters just as much. A listing updated in September 2026 says more about today’s seller expectations than a card that has remained online for three years. When a target property sits well above or below recent peers, the next question is what explains the difference: land, construction quality, project status, furnishing, view, access, title position or simply the seller’s ask. Falling inside the observed range does not make a property fairly valued by itself.

For landed homes, read area in two separate lines: building and land. An apartment comparison will usually lean on internal floor area, while a villa’s total price may be driven materially by its plot. If the source does not make the denominator clear, total price and physical specifications are safer comparison tools than a derived $/m² figure.

This snapshot is therefore an early-stage screening tool. It can narrow the field and sharpen the questions to ask about price, but it does not establish title, boundaries, access, utilities, mandatory costs, building condition or contract terms. Those are property-specific checks better handled through the existing Kampot and Kep guide and the documents for the actual address.