USD 622.20 a month, ex-rent, single person, Numbeo’s August 2026 read on Phnom Penh. Add a centre-of-city 1-BR at USD 617.50 and you are at roughly USD 1,240 before insurance, before a flight to Bangkok for anything serious, before the dollar floor wears through. The brochure number is real. It is also incomplete.
Numbeo’s own comparison puts Bangkok’s ex-rent basket at 1.176× Phnom Penh’s and its centre rent at 1.338× — 17.6% and 33.8% above. The companion comparisons run the other way: Ho Chi Minh City’s ex-rent basket is 0.789× Phnom Penh’s and Manila’s is 0.923×, the 21.1% and 7.7% below that Numbeo prints. Each differential multiplies the Phnom Penh figure; none of them is read backwards off the other city’s. The pensioner pitch (“Cambodia is the cheap alternative”) only holds against Bangkok. Against the other two Western-pensioner candidates in the region, it does not.
The four-city ex-rent matrix
| City | Ex-rent /mo (USD) | 1-BR centre (USD) | Pensioner read |
|---|---|---|---|
| Phnom Penh | Ex-rent /mo (USD) 622.20 | 1-BR centre (USD) 617.50 | Pensioner read The middle of the pack, not the floor. |
| Bangkok | Ex-rent /mo (USD) ~732 | 1-BR centre (USD) ~826 | Pensioner read About 18% more on living, 34% more on rent. |
| Ho Chi Minh City | Ex-rent /mo (USD) ~491 | 1-BR centre (USD) ~577 | Pensioner read About 21% cheaper than PP on living. |
| Manila | Ex-rent /mo (USD) ~574 | 1-BR centre (USD) 540.84 | Pensioner read About 8% cheaper than PP on living. |
Source: Numbeo — Cost of Living comparisons (PP vs BKK, PP vs HCMC, PP vs MNL) · checked 2026-08
Every figure other than Phnom Penh’s two and Manila’s rent is arithmetic on one base: the Phnom Penh value, multiplied by the differential in the matching bilateral comparison. Numbeo computes “X% higher/lower than” as the ratio of the two city indices minus one, so the differential scales the base city’s figure up or down and is never inverted out of the other city’s. Bangkok’s ~732 is 622.20 × 1.176; its ~826 is 617.50 × 1.338. Ho Chi Minh City’s ~491 is 622.20 × 0.789 and its ~577 is 617.50 × 0.935; Manila’s ~574 is 622.20 × 0.923. None of the four cities buys you a private-room ICU bed at a price you would describe in writing. The cheap one is the one that does not.
The visa that is not a visa
Cambodia has no statutory retirement visa. The product foreign agents and Western blogs call “the retirement visa” is an Extension of Stay, Retirement category: an “ER” extension granted on an Ordinary E-class visa to applicants 55 and over. The Ministry of Foreign Affairs and International Cooperation publishes the durations (1, 3, 6, 12 months) and the eligibility shell (55+, Ordinary visa, documentary evidence). It publishes no statutory income threshold and no insurance mandate.
The “USD 800 to USD 1,500 per month” figure that appears in every English-language summary is what officers and agents have asked for in practice. It is not law. There is therefore no published appeal mechanism if an extension is denied, no published premium-loading rule for prior medical history, and no statutory contract to point at when the rules change. This is a feature of the system — not a bug. The regime depends on agent intermediation.
It does include things the Thai O-A and Philippine SRRV do not: no insurance requirement on the applicant, no fixed minimum deposit, no mandatory Cambodian-bank balance to renew. Those are the trade-offs. You have nothing to hold the system to either.
For the comparison the queue actually faces (Thailand O-A versus Philippines SRRV versus this), see the Thai retirement-visa income math, the Philippine SRRV by age and pension, and the rest of the visa-income files.
Dollarised, not dollar
Foreign-currency deposits make up roughly 84% of Cambodia’s broad money; loans and deposits run over 80% in USD; the riel is managed near 4,100 to the USD by the National Bank of Cambodia. A Western pensioner can in principle keep balances, pay rent, and settle hospital bills in USD without a riel touching the books.
In principle. The NBC has phased out small USD notes (USD 1, 2, 5) at banks and ATMs to push riel circulation. Civil-service salaries, taxes, and public-service fees are paid or required in riel. The dollarisation is structural but not contractual; the central bank can compress it further at any time, and has signalled it will. The pensioner who priced his life in dollars in 2026 should model the same life in riel at 4,100 (and at a stressed 4,300) and ask whether the budget survives a slow demonetisation. This is the same risk vector that runs through every dollarised periphery in history. It is not exotic. It is reliable.
Phnom Penh’s USD pegging insulates it from the regional baht curve, but not from a structural NBC move.
Medical evacuation as the standing plan
Phnom Penh’s flagship Western-grade facility is Royal Phnom Penh Hospital, a 100-bed BDMS-group hospital with a 14-bed ICU, 5 operating theatres, and a 24/7 emergency department. Its published emergency service includes Air Evacuation Emergency Medical, an in-house team escorting patients to Bangkok or onward.
Read what that sentence is doing. The best Western-grade hospital in the country advertises evacuation to Bangkok as a standing service. Cardiac catheterisation, complex oncology, major neurosurgery, and high-acuity intensive care are routinely cases for Bangkok (about one hour by air) or Singapore (about two). Phnom Penh stabilises and ships.
The cost structure splits in two. The routine bill is low — primary care, dengue treatment, broken-bone repair, basic surgery. The catastrophic bill is the routine bill plus a charter flight, and the charter is the half nobody prices in advance: local catastrophic-event costs are private and not publicly aggregated, so the figure arrives after the event or not at all. The inversion with Thailand is exact: the Thai O-A makes the policy a condition of the visa in the country a Cambodian resident would be flown to, while Cambodia’s ER asks for no policy at all in the country he would be flown from.
For the broader regional medical comparison, see the cost of aging in Bangkok and the cost of aging in Chiang Mai; for the underwriting collapse at 70, the insurance cliff at 70 and age-out by insurer.
The detail the brochures do not mention
The US Social Security Administration restricts payments to Cambodia.
Cambodia is on SSA’s list of countries to which the Treasury Department prohibits direct payments. US citizens collect Social Security benefits in person at the US Embassy in Phnom Penh, administered through the Federal Benefits Unit at Embassy Manila. Each month the embassy schedules pickup; each month the beneficiary appears. It is a logistical contract with the building.
This works until it does not. Illness, immobility, hospitalisation, evacuation, an embassy closure, a deterioration that prevents in-person attendance: each breaks the payment cycle. The income line is therefore only as durable as the beneficiary’s ability to walk into one specific building on one specific day of the month, and the paperwork that governs what happens when that ends is drafted years earlier or not at all — see the incapacity paperwork trap for what that looks like after the stroke rather than before it.
The UK State Pension is frozen for Cambodia residents; Cambodia is not on the UK reciprocal-uprating list. Australia has no social-security agreement with Cambodia. The pensioner who modelled three lines of income (US SS, UK SP, AUS Age Pension) receives all three at face value in Cambodia, but the US line requires monthly physical presence at a specific Phnom Penh building, the UK line is frozen at first-paid level for life, and the AUS line lacks treaty portability that funds supplement uprating. The combined effect is not catastrophic. It is also not what the brochure suggests.
The border-conflict overlay
A 2025 Thailand-Cambodia border conflict closed the land crossings; a second ceasefire was signed on 27 December 2025. The ceasefire is holding — the two border coordination agencies restated their commitment to strict implementation — and the crossings are still shut. At 30 July 2026, eight months on, they were operating only to send each country’s own nationals home. The Asian Development Bank cut Cambodia’s 2025 growth forecast from 6.1% to 4.9% and the 2026 forecast from 6.2% to 5.0% in September 2025, then cut 2026 again to 4.5% in April 2026 and to 4.1% in July, the last cut of 0.4 percentage points attributed to adverse geopolitical shocks and the prolonged closure of the Thai border. The impact lands through trade, tourism, and the labour corridor. The ceasefire holding and the border reopening are separate facts, and only the first has happened.
The standing medical-evacuation plan runs through the air corridor either way; commercial flights and dedicated medical aircraft cross Cambodia-Thai airspace on routes that were not closed during the land-border conflict. The reliability question is structural: a region where the two principal medical-evacuation destinations (Bangkok, Singapore) are in different countries from the residence, and where one of those bilateral relationships is intermittently kinetic, carries a different risk profile from a region where it is not. The pensioner who built his life around evacuation-to-Bangkok has one destination in the plan and one bilateral relationship underneath it. Whether a given policy carries a second evacuation destination at all is a question its schedule answers and the brochure does not.
The banking sector at confidence-test level
The Cambodian banking sector’s June 2025 non-performing loan ratio sat at 8.1 to 8.3%, roughly USD 4.7 billion in NPLs, the highest in a decade. The loan-to-deposit ratio improved to 95% for the first time in five years; Cambodia Investment Review, in an April 2026 sector piece, described the moment as “a confidence test, not a crisis”, explicitly rebutting “rumours of US action”.
Separately, and on a different regime entirely, the documentation is tightening. ABA Bank’s December 2025 General Terms classify uncooperative US persons as recalcitrant under FATCA, requiring W-9 or W-8BEN documentation — the same FATCA / CRS paperwork standard now spreading through SE Asian banking everywhere, which means Cambodia is not singling anyone out and there is nowhere in the region to move it to.
The pensioner-relevant exposure is concentration, and it is arithmetic rather than prediction. Operating cash sitting inside a single dollarised banking sector, in its highest-NPL year in a decade, with no leg outside it, is a position whose value depends on depositor confidence in that dollarisation holding — which is the precise thing the sector is currently being asked about in public. Nothing in the published figures says it will not hold. They also do not say what the position is worth if it does not.
Air, infection, road
Phnom Penh’s 2026 PM2.5 reading sat at 18 µg/m³ on IQAir’s March 2026 newsroom data (about 3.6× the WHO 2021 guideline of 5 µg/m³), with IQAir placing the city in its top-10 most-polluted-major-cities tranche for that period. For an older respiratory or cardiovascular patient that delta is not cosmetic; the chronic exposure shifts mortality, slowly, in the direction the brochures avoid.
Cambodia recorded 18,987 dengue cases and 46 deaths in 2024, down 46% year on year; the first half of 2025 saw about 7,000 cases and 15 deaths, all among children. Adult mortality is low but not zero; the chronic-condition pensioner whose immune reserve is already compromised carries a different curve than the median statistic.
Road traffic killed 1,509 people in Cambodia in 2024 (about 4.1 per day, down 5% year on year) at a fatality rate of roughly 18.8 per 100,000 against the Asia-Pacific average of 15.2 and the SE Asia average of 14.4. The pensioner read is the same as everywhere in the region: the foreign pensioner on a motorbike is the part of the curve insurers exclude. See the road death multiplier for foreign pensioners.
Property as the asset trap
Foreigners cannot own land in Cambodia. The available structures are strata-title condominium ownership (up to 70% of units in a building, above ground floor, more than 30 km from a national land border) and leasehold (50 years renewable). The strata regime is real but the resale market is thin; the 2026 BKK1-area condominium market ran USD 2,650 to USD 3,500 per square metre, with the city average projected near USD 2,783 per square metre.
The trap is not the price. It is the exit. A 70-year-old who buys a USD 200,000 BKK1 unit in 2026 must, at 80, sell it back into a market with the same 70% foreign cap, the same thin buyer pool, and the same agent-discretionary closing process, at whatever the market is then. It is the thinness of that exit that traps him; the entry price told him nothing about it. Phnom Penh is the more illiquid case.
What the cheap headline obscures
Phnom Penh is cheaper than Bangkok and more expensive than Manila and Ho Chi Minh City on the same Numbeo basket. The retirement-visa contract is informal and agent-discretionary. The medical-evacuation plan is a standing monthly line item, budgeted like rent. The US Social Security cheque is collected at a building. The banking sector is in its highest-NPL year in a decade and rebuts sanctions rumours in public. The regional bilateral relationship that hosts the primary medical-evacuation destination ceased fire on 27 December 2025 and has kept its land border shut to ordinary traffic ever since. Each of these is verifiable. Together they do not describe a cheap retirement — they describe a cheap headline.
There is nothing wrong with choosing Phnom Penh anyway. The wrong move is choosing it on the headline and learning the rest at 75, after the cataract surgery, in the embassy queue, on a riel-denominated lease. The data does not stop being the data when you are tired.