A cheap house is not necessarily a good investment. Neither is an expensive apartment with an impressive advertised rent. The question is whether a particular property can attract a tenant, pay its full costs and still leave its owner with a return that justifies the work and risk.
Short answer: For a conventional long-term rental, a T2 apartment in Vila Nova de Gaia offers the best balance of rental economics and depth of observed transactions in this screen. Torres Vedras houses deserve a closer look for buyers prepared to maintain a house. Setúbal T3 apartments have a higher headline rent-to-price ratio, but the rented homes in the sample are materially larger than those sold. Sines and Santiago do Cacém have the largest apparent ratios and the thinnest evidence; treat them as research leads, not proven yields.
We compared municipal used-home sale prices with advertised long-term rental asking rents for the same broad property type, then checked the prior quarter. This covers mainland Portugal; the data available for this analysis do not establish a comparable shortlist for Madeira or the Azores. It does not measure holiday-let occupancy or returns.
Crucial distinction: an asking rent is what a landlord seeks, not what a tenant ultimately agrees to pay. The figures below compare two different sets of homes. Even within the same bedroom category, location, condition, parking and floor area can differ. The percentages are screening tools, not achieved gross yields, forecasts or net returns.
The national screen: high ratios are not equally reliable.
Annualized asking rent ÷ average used-home sale price, Q2 2026. Labels show the number of sales and rental offers observed. Color flags how we use each market in this article; it does not measure investment quality.
Accessible data table
| Market | Observed counts | Ratio | Use class |
|---|---|---|---|
| Sines apartments | 12 sales · 23 offers | 10.28% | exploratory |
| Alcochete apartments | 42 sales · 26 offers | 9.41% | exploratory |
| Santiago do Cacém apartments | 30 sales · 8 offers | 9.35% | exploratory |
| Viana do Castelo houses | 87 sales · 16 offers | 9.04% | exploratory |
| Setúbal T3 apartments | 144 sales · 36 offers | 7.23% | comparison |
| Torres Vedras houses | 180 sales · 17 offers | 6.90% | priority |
| Gaia T3 apartments | 264 sales · 64 offers | 6.64% | priority |
| Gaia T2 apartments | 426 sales · 126 offers | 6.29% | priority |
| Caldas da Rainha apartments | 149 sales · 15 offers | 5.97% | comparison |
| Évora houses | 105 sales · 31 offers | 5.48% | comparison |
| Braga T2 apartments | 86 sales · 45 offers | 5.45% | comparison |
| Figueira da Foz apartments | 191 sales · 23 offers | 5.43% | comparison |
Source data: Confidencial Imobiliário, SIR, Q2 2026. Calculations: The American’s Guide to Portugal. An asking-rent screen excludes vacancies, transaction costs, repairs, taxes, financing and sale costs. Santiago do Cacém has fewer than ten rental offers and is included only to show why the most eye-catching figure deserves investigation.
How we narrowed
The underlying Q2 sales dataset contains observations for 272 named municipalities; the asking-rent dataset contains observations for 171. We paired used-home sales with rental offers by municipality and broad housing type. The primary comparison required at least 30 sales, 10 rental offers and average sale and rental floor areas within 20% of each other, leaving 145 comparable municipality/property-type combinations. We also checked Q1 counts, prices, rents and sizes. Those cutoffs are editorial safeguards, not statistical confidence intervals.
The average sale is not the average rental home, and neither is necessarily available to buy today. We favor a repeatable signal, plausible local tenancy and a property an investor can underwrite over a single spectacular percentage. Numbers suppressed or absent in the source are left absent; we do not estimate a yield from them.
Why the prior quarter matters: One quarter can capture a few unusual homes. Consistency helps identify candidates for further work, although similar numbers across two quarters do not guarantee future rent or price performance.
Did the apparent rent-to-price relationship persist?
Two-quarter comparison of asking-rent screens. The six pairs shown use the same property type in each municipality. Blue is Q1; dark teal is Q2.
Accessible Q1/Q2 table
| Market | Q1 | Q2 |
|---|---|---|
| Santiago do Cacém apartments | 9.45% | 9.35% |
| Viana do Castelo houses | 8.71% | 9.04% |
| Torres Vedras houses | 7.03% | 6.90% |
| Gaia T2 apartments | 6.42% | 6.29% |
| Évora houses | 5.52% | 5.48% |
| Braga T2 apartments | 6.02% | 5.45% |
Source data: Confidencial Imobiliário, SIR, Q1–Q2 2026. Calculations: The American’s Guide to Portugal. Small or changing samples can produce similar ratios for different underlying reasons; see counts and caveats in the text.
Where I would start looking
- Vila Nova de Gaia: the most useful repeatable apartment screen. A used T2 apartment averaged €257,438 across 426 Q2 sales. The corresponding asking rent averaged €1,349 across 126 offers. The implied asking-rent screen is 6.29%, against 6.42% in Q1; average floor areas were 95 m² sold and 100 m² offered. This is a far deeper and more closely matched comparison than many high-ratio inland markets. A Gaia T3 screens at 6.64% (€311,036 against €1,721 asked, with 264 sales and 64 rental offers), but requires more capital. Start with T2s when budget and ease of reletting matter; compare an individual T3 when the extra bedroom earns enough additional rent to cover its purchase and running costs. The municipality covers varied neighborhoods: evaluate actual commute, building condition, condominium finances and nearby competing rentals rather than treating Gaia as one uniform submarket.
- Torres Vedras: an unexpectedly strong house market. Houses averaged €300,618 across 180 sales and €1,729 per month across 17 rental offers, a 6.90% asking-rent screen. Q1 was 7.03%, and the Q2 average sizes almost coincide: 144 m² sold and 142 m² offered. This is the strongest overlooked mainland comparison for a buyer willing to do more property-level work. That work matters. A municipality-wide house average may blend very different villages, beach areas, plots and building conditions. Obtain comparable year-round leases for the precise location. Inspect the roof, damp, heating, energy certificate, permits, exterior works and access. A lower-maintenance apartment may outperform a higher-headline-yield house after repairs.
- Setúbal: attractive numbers, with an important size warning. Q2 used T3 apartments averaged €244,749 over 144 sales; rental offers averaged €1,475 over 36, implying 7.23%. Q1 screened at 6.90%. Yet Q2 sold T3s averaged 100 m² while offered T3s averaged 118 m². That gap can make the simple calculation flattering. Seek a signed-rent comparison for a unit of similar area, condition and location before relying on €1,475. Setúbal's all-apartment category has more observations—386 sales and 84 rental offers—but combines bedroom sizes and still has an 11% floor-area difference. It is a useful second check, not a substitute for the precise T3 comp.
- The Alentejo coast: the biggest lead, and the biggest verification task. Sines apartments averaged €254,000 across only 12 used-home sales and €2,175 across 23 rental offers in Q2: a 10.28% screen. The source does not publish a comparable Q1 sale average for this segment. Nearby Santiago do Cacém apartments screened at 9.35% (€231,859 against €1,806), close to 9.45% in Q1, but Q2 had 30 sales and only eight rental offers. These figures are too tempting to ignore and too thin to call investable returns. Sines has a port, industry and announced investment projects, plausible sources of rental interest. It would be a mistake to convert those announcements directly into a guaranteed permanent-tenant rent. Confirm whether the observed offers target long-term households, temporary project workers or furnished corporate stays; check actual lease terms, local employment dependence and what happens if a major project ends. AICEP describes the region's port and industrial base.
- Viana do Castelo: house economics that need a repair reserve. Used houses averaged €246,769 over 87 sales; asking rents averaged €1,858 over 16 offers. That produces 9.04%, versus 8.71% in Q1. But the earlier rental sample had only nine offers, and average house sizes changed materially between quarters. I would investigate this as a specific-house opportunity, not promise a 9% rental yield. Document heating, moisture, structural condition, tenant demand and the cost of grounds or exterior maintenance.
Évora houses — Alentejo · steadier comparison
€248,560 average used price; €1,135 average asking rent; 105 sales and 31 offers. The screen held near 5.5% in both quarters. Useful for a buyer prioritizing consistent observations over a very high headline ratio.
Caldas da Rainha apartments — Silver Coast · lower entry point
€213,000 used price; €1,059 asked rent; 5.97%, close to 5.92% in Q1. Rental offers averaged 105 m² versus 92 m² sold, so confirm a like-for-like unit. Figueira da Foz apartments offer a lower €182,500 average entry at a 5.43% screen.
Braga T2 apartments — Northern Portugal · conventional apartment
€230,997 used price; €1,050 asked rent; 86 sales and 45 offers. The 5.45% Q2 screen was 6.02% in Q1. Sale and rental sizes were close. It belongs on an apartment shortlist, but purchase-price discipline matters.
Portimão and Faro — Algarve · year-round lens
Portimão used apartments screened at 6.22% (€263,036 / €1,363) with 360 sales and 42 offers. Faro apartments screened at 6.87% (€326,947 / €1,872) with 107 sales and 62 offers; Faro's sale mix changed enough that its Q1 8.22% should not be read as a trend. Neither screen assumes holiday-let income.
Why cheapest places did not top the list
Guarda T3 apartments averaged €154,000 in Q2, but €595 asking rent gives a 4.64% screen. Viseu apartments screened at 4.11%, Coimbra apartments at 4.15% and Aveiro apartments at 4.34%. Those are not bad places to live; their observed rents simply did not rise enough relative to the purchase prices to lead this particular long-term rental comparison.
The Serra da Estrela deserves a separate local study. Covilhã had 88 apartment sales observations but only six apartment rental offers in Q2, with no published average asking rent for a sound like-for-like comparison. The source similarly gives too little rental evidence for Beja and Portalegre, despite sale transactions. Do not infer that demand is absent: thin recorded offers can reflect the data source, a small or differently marketed rental sector, or both. These places remain research candidates, not ranked investment recommendations.
A surprising Greater Lisbon lead: Alcochete apartments screened at 9.41% in Q2 and 8.50% in Q1. But Q2 rented apartments averaged 107 m² against 128 m² for those sold, and the municipality-level mix needs inspection. Put it on a transaction-level research list rather than placing the headline percentage above Gaia's more abundant observations.
Actual purchase: what happens when you underwrite one?
Consider a purely illustrative Torres Vedras house purchased for the Q2 used-house average, €300,618, and rented at the Q2 average asking rent, €1,729 per month. An average is not a listing and an asking rent is not a signed lease; this is a sensitivity exercise, not a property forecast.
From advertised rent to before-tax cash flow.
Illustration only: one empty month; operating expense allowance of 20% of collected rent; 70% mortgage financing at 4% nominal interest over 30 years. The 20% allowance is a planning assumption, not measured local operating costs.
- 12 months at advertised rent€20,748
- Less one empty month−€1,729
- Collected rent€19,019
- Less 20% operating allowance−€3,804
- Income before debt and tax€15,215
- Less annual mortgage payments−€12,056
- Before-tax cash flow≈€3,160
Upfront cash illustration ≈€111,229: 30% down (€90,185) plus a provisional 7% acquisition-cost allowance (€21,043). Excludes immediate repairs and a cash reserve.
Before-tax cash-on-cash: about 2.84%. A headline 6.90% asking-rent screen becomes much smaller after one empty month, operating costs, financing and upfront costs. This is not the owner's after-tax return.
The 7% acquisition allowance is an illustrative placeholder, not a claim about the applicable IMT rate. Calculate actual IMT for the buyer and property, plus stamp duty, legal and registration charges, loan fees and any renovation. The property's VPT and intended use can affect the tax base or rate.
Purchase price is a decision, not a footnote. Hold that illustrative €1,729 monthly asking rent and all the assumptions above constant. At €240,000, annual before-tax cash flow would be about €5,591; at €300,618, €3,160; at €375,000, only €177; and at €450,000, a loss of roughly €2,831. The €375,000 example is almost break-even before income tax, major repairs and eventual selling costs. Paying above the municipal average for the right property may still make sense, but the expected rent needs independent evidence.
Appreciation and exit: three ways the same house might finish. Suppose that €300,618 example is sold after ten years, with 5% illustrative selling expenses. At 0% annual price growth, sale proceeds after expenses and the remaining illustrative mortgage balance are about €119,800; at 2% growth, about €182,342; and at 5% growth, about €299,404. Against the €111,229 initial down payment plus provisional acquisition allowance, those differences are approximately €8,571, €71,113 and €188,175, respectively. They exclude ten years of rental cash flow, income and capital-gains taxes, capital improvements, possible prepayment charges, foreign-exchange movements and the cost of reserves. The high-growth case is a scenario, not a prediction. Check who would buy the home at exit and how long similar homes take to sell.
Taxes can reverse the ranking. Portugal's 2026 framework provides a 10% autonomous IRS rate for qualifying residential rental income within a moderate-rent limit, through 2029, subject to legal conditions. The law ties the 2026 mainland limit to 2.5 times the €920 minimum monthly wage: €2,300 per month. An asking rent below that ceiling alone does not establish eligibility. The actual IMT treatment also depends on the buyer, intended use and applicable exceptions. A U.S. taxpayer must additionally model U.S. rental reporting, depreciation and potential foreign tax credits with a cross-border adviser. Do not subtract a flat 10% from gross rent and call the remainder profit.
How to use this guide when shopping
- Pick a tenant and property type. For a standard apartment, begin with Gaia T2s and compare Braga or Setúbal units. For houses, compare Torres Vedras and Viana do Castelo with a realistic repairs budget. Treat the Alentejo coast as a specialist research project.
- Build an address-level rent comp. Obtain at least three comparable year-round listings and, where possible, signed leases. Match floor area, bedrooms, condition, parking, furnishings and exact neighborhood. Discount asking rents when the evidence requires it.
- Price the purchase in full. Calculate IMT and stamp duty for the actual transaction; add legal, registry, lender, insurance and inspection costs, repairs and a reserve. Ask the condominium for budgets and planned works if buying an apartment.
- Stress-test it. Try lower rent, two empty months, a major repair, refinancing at a higher rate and 0% appreciation. Estimate after-tax results in Portugal and the United States separately before combining them.
- Plan the exit before signing. Identify likely future buyers, competing supply, likely selling costs and the condition required for resale. A paper gain is not an exit strategy.
FAQ
- What is the best city in Portugal to buy an investment property?
- For an investor seeking a conventional long-term apartment rental, Vila Nova de Gaia's used T2 segment offers the strongest blend of observed sale and rental depth, relatively comparable home sizes and a steady Q1–Q2 asking-rent screen in this analysis. It does not follow that any specific Gaia T2 is a good purchase.
- Where are rental yields highest in Portugal?
- Our highest apparent Q2 asking-rent-to-sale-price comparison was Sines apartments, followed by Alcochete apartments and Santiago do Cacém apartments. These are not measured rental yields; sample size, property mix, signed rent, expenses and acquisition costs must be checked before drawing an investment conclusion.
- Is the Silver Coast a better investment than Lisbon?
- Some Silver Coast areas offer a lower entry price. Caldas da Rainha apartments screened at 5.97% and Figueira da Foz apartments at 5.43% in Q2. That can be attractive for a particular buyer, but it is not proof of stronger net returns or an easier exit than a specific Lisbon-area property.
- Should Americans buy a holiday rental in the Algarve?
- This guide measures a long-term rental screen and cannot answer that question. A holiday rental requires separate evidence for occupancy, nightly rates, management, seasonality, local licensing and applicable rules. Portimão and Faro appear here only as year-round rental comparisons.
- Does a rent below €2,300 automatically qualify for Portugal’s 10% rental tax treatment?
- No. The 2026 mainland rent ceiling is one condition in a legal regime with other requirements and a stated end date. Ask a qualified Portuguese tax adviser to confirm the contract and ownership facts, and a U.S. tax adviser to calculate the investor’s American obligations.
What to do next
Use the markets above to choose where to investigate, then model an individual property using its actual purchase terms and verifiable rent. If you are comparing homes across Portugal, the quality of that property-level evidence matters more than winning a national ranking.
Methodology and attribution
Source data: Confidencial Imobiliário, SIR, Q1–Q2 2026. Calculations and editorial analysis: The American’s Guide to Portugal. © IMOESTATISTICA – TODOS OS DIREITOS RESERVADOS. Used-home sale means and advertised rental means come from different sets of properties, and counts reflect the observations in those tables. Data are municipal, not neighborhood-level; missing or suppressed source averages were not estimated. This article provides bounded comparisons, not a reproduction of the underlying national dataset. Last fact-check of the legal links: 21 September 2026.
- Decree-Law 97/2026 https://diariodarepublica.pt/dr/detalhe/decreto-lei/97-2026-1124493227 and Tax Authority, EBF Article 45-C https://info.portaldasfinancas.gov.pt/pt/informacao_fiscal/codigos_tributarios/bf_rep/Pages/ebf45c.aspx.
- Decree-Law 139/2025 https://diariodarepublica.pt/dr/detalhe/decreto-lei/139-2025-992879809.
- Portuguese Tax Authority, purchase taxes https://info.portaldasfinancas.gov.pt/pt/apoio_ao_contribuinte/Cidadaos/Casa_e_propriedades/Compra_da_casa/Paginas/default.aspx and IMT Article 17 https://info.portaldasfinancas.gov.pt/pt/informacao_fiscal/codigos_tributarios/cimt/Pages/cimt17.aspx.
- IRS Publication 527 https://www.irs.gov/publications/p527 and IRS Publication 514 https://www.irs.gov/publications/p514.
- AICEP, Alentejo Litoral regional investment profile https://investinportugal.portugalglobal.pt/regions/alentejo-litoral/.
.jpg){kind=link}